Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks441 from public data
- YOLOv6: A Single-Stage Object Detection Framework for Industrial Applications3,433
This technical report strives to push the YOLO series to the next level, stepping forward with an unwavering mindset for industry application, and builds a suite of deployment-ready networks at various scales to accommodate diversified use cases.
- Towards Discriminability and Diversity: Batch Nuclear-Norm Maximization Under Label Insufficient Situations454
To improve both discriminability and diversity, the proposed Batch Nuclear-norm Maximization (BNM) on the output matrix could boost the learning under typical label insufficient learning scenarios, such as semi-supervised learning, domain adaptation and open domain recognition.
- 350
- Age-Invariant Face Recognition by Multi-Feature Fusionand Decomposition with Self-attention238
A novel Multi-feature Fusion and Decomposition framework for age-invariant face recognition, which learns more discriminative and robust features and reduces the intra-class variants and improves the AIFR performance than previous state-of-the-art methods.
- Parsing-Based View-Aware Embedding Network for Vehicle Re-Identification229
This paper proposes a parsing-based view-aware embedding network (PVEN), which helps capture the stable discriminative information of vehicle under different views and design a common-visible attention to focus on the common visible views, which not only shortens the distance among intra-instances, but also enlarges the discrepancy of inter-instance.
- A Self-Attentive Model with Gate Mechanism for Spoken Language Understanding171
A novel self-attentive model with gate mechanism to fully utilize the semantic correlation between slot and intent and outperforms other popular methods by a large margin in terms of both intent detection error rate and slot filling F1-score is proposed.
- FaceVerse: a Fine-grained and Detail-controllable 3D Face Morphable Model from a Hybrid Dataset157
FaceVerse is presented, a fine-grained 3D Neural Face Model, which is built from hybrid East Asian face datasets containing 60K fused RGB-D images and 2K high-fidelity 3D head scan models and a novel coarse-to-fine structure is proposed to take better advantage of the hybrid dataset.
- State-Relabeling Adversarial Active Learning150
This paper proposes a state relabeling adversarial active learning model (SRAAL), that leverages both the annotation and the labeled/unlabeled state information for deriving the most informative unlabeled samples.
- Dependency Exploitation: A Unified CNN-RNN Approach for Visual Emotion Recognition135
This paper proposes a unified CNNRNN model to predict the emotion based on the fused features from different levels by exploiting the dependency among them and leverages convolutional neural network with multiple layers to extract different levels of features within a multi-task learning framework.
- A Novel Augmented Reality Navigation System for Endoscopic Sinus and Skull Base Surgery: A Feasibility Study125
It is concluded that this navigational system will provide rhinologists with more intuitive and more detailed imaging information, thus reducing the judgment time and mental workload of surgeons when performing complex sinus and skull base surgeries.
- Open-Vocabulary One-Stage Detection with Hierarchical Visual-Language Knowledge Distillation124
This paper proposes a hierarchical visual-language knowledge distillation method, i.e., Hi-erKD, for open-vocabulary one-stage detection, which significantly surpasses the previous best one- stage detector, and reduces the AP50 performance gap from 14% to 7.3% compared to the best two-stage detector.
- Parallel deblocking filter for HEVC on many‐core processor123
Experiments show that the proposed three-step parallel framework for HEVC dramatically accelerates more than the state-of-the-art parallel method.
- 114
- 108
- Let the Data Choose: Flexible and Diverse Anchor Graph Fusion for Scalable Multi-View Clustering101
A novel anchor-based method termed Flexible and Diverse Anchor Graph Fusion for Scalable Multi-view Clustering (FDAGF) and a novel hybrid fusion strategy for multi-size anchor graphs with theoretical proof, which allows flexible and diverse anchor graph fusion.
- Adaptive Reconstruction Network for Weakly Supervised Referring Expression Grounding98
A novel end-to-end adaptive reconstruction network (ARN) that builds the correspondence between image region proposal and query in an adaptive manner: adaptive grounding and collaborative reconstruction.
- Few Shot Generative Model Adaption via Relaxed Spatial Structural Alignment96
This work proposes a relaxed spatial structural alignment (RSSA) method to calibrate the target generative models during the adaption and designs a cross-domain spatial structural consistency loss comprising the self-correlation and disturbance correlation consistency loss.
- 91
- 91
- Efficient parallel HEVC intra‐prediction on many‐core processor90
Experiments show that the proposed method dramatically accelerates more than the state-of-the-art parallel method for HEVC IP.
- Real-World Person Re-Identification via Degradation Invariance Learning84
By introducing a self-supervised disentangled representation learning strategy, this framework is able to simultaneously extract identity-related robust features and remove real-world degradations without extra supervision and is ableTo achieve state-of-the-art performance on several Re-ID benchmarks.
- Symbolic Music Generation with Transformer-GANs83
It is demonstrated via human evaluations and a new discriminative metric that the music generated by the approach outperforms a baseline trained with likelihood maximization, the state-of-the-art Music Transformer, and other GANs used for sequence generation.
- Detection of microplastics in domestic and fetal pigs’ lung tissue in natural environment: A preliminary study75
The presence of MPs in the lung tissue of both domestic and fetal pigs in the natural environment was confirmed, but the main characteristics differed, which indicated the increasing risk of MPs to human respiratory tract is increasing.
- 75
- 72
- From Speaker to Dubber: Movie Dubbing with Prosody and Duration Consistency Learning71
This work proposes a two-stage dubbing method that allows the model to first learn pronunciation knowledge before practicing it in movie dubbing, and outperforms several state-of-the-art methods on two primary benchmarks.
- Knowledge-guided Pairwise Reconstruction Network for Weakly Supervised Referring Expression Grounding69
A knowledge-guided pairwise reconstruction network (KPRN) is proposed, which models the relationship between the target entity (subject) and contextual entity (object) as well as grounds these two entities.
- Online Video Understanding: OVBench and VideoChat-Online67
OVBench, a comprehensive question-answering benchmark designed to evaluate models’ ability to perceive, memorize, and reason within online video contexts, is introduced and VideoChat-Online, a robust and efficient model for online video understanding is developed.
- Rethinking Graph Neural Architecture Search from Message-passing65
The proposed Graph Neural Architecture Search (GNAS) can automatically learn better architecture with the optimal depth of message passing on the graph and achieves remarkable improvement over state-of-the-art manual designed and search-based GNNs on five large-scale datasets at three classical graph tasks.
- Toward Energy-Efficient Federated Learning Over 5G+ Mobile Devices64
This article investigates how to develop energy- efficient FL over 5G+ mobile devices by making a trade-off between energy consumption for “working” and “talking” in order to boost the overall energy efficiency.
- Retracted: Efficiency Analysis of Machine Learning Intelligent Investment Based on K-Means Algorithm64
The results show that the machine learning based on K-means algorithm makes a concrete analysis of the investment efficiency of Capricorn Smart Investment, and this method can also be used for the efficiency analysis of other smart investment products.
- 64
- 64
- Assessment model for perceived visual complexity of painting images61
The results show that the regression model has a good ability of measuring aesthetic quality, beauty, and liking of color of the painting images involved in JenAesthetics dataset and the proposed classification method can predict the visual complexity perception of paintings with an accuracy of 86.78%.
- 61
- 58
- EvoPass: Evolvable graphical password against shoulder-surfing attacks56
Two metrics, Information Retention Rate (IRR) and Password Diversity Score (PDS) are introduced to guide the process of generating pass sketches and a set of challenge images to improve password strength gradually over time in EvoPass.
- 56
- PDBL: Improving Histopathological Tissue Classification With Plug-and-Play Pyramidal Deep-Broad Learning55
Experimental results demonstrate the proposed PDBL can steadily improve the tissue-level classification performance for any CNN backbones, especially for the lightweight models when given a small among of training samples (less than 10%).
- Semantic Relation-aware Difference Representation Learning for Change Captioning50
A semantic relation-aware difference representation learning network to explicitly learn the difference representation in the existence of distractors is proposed and an attention-based visual switch is devised to dynamically use visual information for caption generation.
- LiDAR-PTQ: Post-Training Quantization for Point Cloud 3D Object Detection48
An effective PTQ method called LiDAR-PTQ, which is particularly curated for 3D lidar detection (both SPConv-based and SPConv-free) and is cost-effective being $30\times faster than the quantization-aware training method.
- Diterpenes from the Hainan Soft Coral Lobophytum cristatum Tixier-Durivault47
The present work supports Faulkner's proposal of prenylgermacrene as the precursor of many diterpenes and the absolute configurations of lobophytumins A, B, and D-F were also tentatively suggested.
- EvalMuse-40K: A Reliable and Fine-Grained Benchmark with Comprehensive Human Annotations for Text-to-Image Generation Model Evaluation46
Two new methods to evaluate the image-text alignment capabilities of T2I models are introduced: FGA-BLIP2 which involves end-to-end fine-tuning of a vision-language model to produce fine-grained image-text alignment scores and PN-VQA which adopts a novel positive-negative VQA manner in VQA models for zero-shot fine-grained evaluation.
- Secure Social Recommendation Based on Secret Sharing44
A SEcure SOcial RECommendation (SeSoRec) framework which can collaboratively mine knowledge from social platform to improve the recommendation performance of the rating platform, and securely keep the raw data of both platforms is proposed and its correctness and security theoretically is proved.
- RestainNet: A self-supervised digital re-stainer for stain normalization42
This paper formulated stain normalization as a digital re-staining process and proposed a self-supervised learning model, which is called RestainNet, which outperforms existing approaches and achieves state-of-the-art performance with regard to color correctness and structure preservation.
- TRANSIENT AUDITORY STORAGE OF ACOUSTIC DETAILS IS ASSOCIATED WITH RELEASE OF SPEECH FROM INFORMATIONAL MASKING IN REVERBERANT CONDITIONS41
The ability to temporally store fine details contributes to perceptual integration of correlated leading and lagging sounds, which in turn contributes to releasing speech from informational masking in noisy, reverberant environments.
- Image classification base on PCA of multi-view deep representation40
This paper proposes an image classification technology using principal component analysis based on multi-view depth characters using the SVM and shows that the method has good performance.
- Attentive Recurrent Neural Network for Weak-supervised Multi-label Image Classification40
An end-to-end attentive recurrent neural network for multi-label image classification under only image-level supervision, which learns the discriminative feature representations and models the label relations simultaneously.
- 38
- Nonlocal Means-Based Denoising for Medical Images36
The proposed adaptive NL-means denoising scheme has been compared with similar state-of-art methods over synthetic and real clinical medical images showing an improved performance in all cases analyzed.
- 36
- R&B: Region and Boundary Aware Zero-shot Grounded Text-to-image Generation35
A Region and Boundary aware cross-attention guidance approach that gradually modulates the attention maps of diffusion model during generative process, and assists the model to synthesize images with high fidelity, and design a region-aware loss to refine the generative layout during diffusion process.
- Rˆ3Net:Relation-embedded Representation Reconstruction Network for Change Captioning34
A Relation-embedded Representation Reconstruction Network (Rˆ3Net) is proposed to explicitly distinguish the real change from the large amount of clutter and irrelevant changes and introduces a syntactic skeleton predictor (SSP) to enhance the semantic interaction between change localization and caption generation.
- Pinning bipartite synchronization for inertial coupled delayed neural networks with signed digraph via non-reduced order method34
The study investigates bipartite synchronization of inertial coupled delayed neural networks (ICDNNs) with signed digraph by non-reduced order method and pinning control and develops a novel Lyapunov-Krasovskii functional which depends on the topology of the ICDNNs.
- A Hierarchical CNN-RNN Approach for Visual Emotion Classification34
A hierarchical convolutional neural network-recurrent neural network approach to predict the emotion based on the fused stimuli by exploiting the dependency among different-level features is proposed.
- 34
- Unifying Structured Data as Graph for Data-to-Text Pre-Training33
This paper proposes a structure-enhanced pre-training method for D2T generation by designing a structure-enhanced Transformer, and devise a position matrix for the Transformer, encoding relative positional information of connected nodes in the input graph.
- 33
- 31
- 29
- 29
- Bidirectional difference locating and semantic consistency reasoning for change captioning28
This work proposes a framework with bidirectional difference localization and semantic consistency reasoning to describe the image changes and shows that the performance of the proposed model outperforms the state‐of‐the‐art change captioning models by a large margin.
- Jointly optimizing global and local color consistency for multiple image mosaicking28
This paper proposes a novel image color correction approach that can correct global and local color discrepancies simultaneously and preserve image gradient as much as possible and demonstrates that the proposed approach outperforms the state-of-the-art color correction approaches in both visual quality and quantitative metrics.
- 28
- 28
- 27
- 26
- Fine-grained Feature Alignment with Part Perspective Transformation for Vehicle ReID26
This paper proposes part perspective transformation module (PPT) to map the different parts of vehicle into a unified perspective respectively and proposes a dynamically batch hard triplet loss to select the common visible regions of the compared vehicles.
- Minimum Width of Leaky-ReLU Neural Networks for Uniform Universal Approximation25
This paper examines a uniform UAP for the function class C (K, R) and gives the exact minimum width of the leaky-ReLU NN as $w_{\min}=\max(d_x,d_y)+\Delta (d_x, d_y)$, where $\Delta (d_x, d_y)$ is the additional dimensions for approximating continuous functions with diffeomorphisms via embedding.
- Learning salient features to prevent model drift for correlation tracking24
This paper proposes a multi-model structure for CF trackers to capture the target appearance changes, where different appearance models are trained with specific samples to catch the salient features of the target and reduce the computational cost.
- 23
- Weakly Supervised Text-based Actor-Action Video Segmentation by Clip-level Multi-instance Learning23
This work is the first to perform AAVSS under weakly supervised situations and achieves comparable performance to state-of-the-art fully supervised approaches on two large-scale benchmarks, including A2D Sentences and J-HMDB Sentences.
- Position-domain integrity risk-based ambiguity validation for the integer bootstrap estimator23
A partial ambiguity resolution method is developed in order to satisfy the predefined integrity risk requirement and it is revealed that the proposed method obtains the best balance between the positioning accuracy and the continuity performance.
- 23
- Research and improvement on Apriori algorithm of association rule mining23
The improved Apriori algorithm improved it from three aspects: the strategy of the join step and the prune step was improved when candidate frequent (k+1)-itemsets were generated from frequent k-itemsets, and the method of dealing with transaction was improved to reduce the time of pattern matching to be used in the A Priori algorithm.
- Distributed image understanding with semantic dictionary and semantic expansion22
This paper proposes a semantic dictionary to describe the images on the level of semantic, which characterizes the probability distribution between visual appearances and semantic concepts, and the learning procedure of semantic dictionary is formulated into a minimization optimization problem.
- 21
- Learning Degradation-Invariant Representation for Robust Real-World Person Re-Identification21
A degradation invariance learning framework for robust person re-identification that is able to continuously maintain degradation-invariant priors from one or more datasets to improve the robustness of identity features, achieving state-of-the-art Re-ID performance on several challenging real-world benchmarks with a unified model.
- 21
- IR-GAN: Image Manipulation with Linguistic Instruction by Increment Reasoning21
An IR-GAN is proposed, which aims to reason the consistency between visual increment in images and semantic increment in instructions, and introduces the word-level and instruction-level instruction encoders to learn user's intention from history-correlated instructions as semantic increment.
- Training Efficient Saliency Prediction Models with Knowledge Distillation21
This paper proposes a condensed and randomly initialized student network employed to achieve higher efficiency by transferring knowledge from complicated and well-trained teacher networks for efficient pixel-wise saliency prediction and proposes two statistical representations of feature maps as knowledge.
- 20
- 20
- 20
- 20
- 20
- Cluster Number Variability Problem in LEACH20
This work proposes an improved clustering scheme (I-LEACH) to correct a slight inaccuracy in the computation of the node self-selected probability per round and shows that this scheme can ensure a more stable number of clusters.
- 19
- Multi-Features Fusion and Decomposition for Age-Invariant Face Recognition19
This work proposes a novel Multi-Features Fusion and Decomposition (MFFD) framework to learn more discriminative feature representations and alleviate the intra-class variations for AIFR.
- Extending Implicit Discourse Relation Recognition to the PDTB-319
This work presents data to support the claim that the PDTB-3 contains many more Implicit discourse relations, and methods that can serve as a non-trivial baseline for future state-of-the-art recognizers for implicit discourse relations.
- Integrity monitoring of high-accuracy GNSS-based attitude determination19
The proposed integrity monitoring method constructs double test statistics to satisfy the requirements of integrity risk and continuity risk, simultaneously and can reduce the maximum attitude error by about 2.3°, when compared with the standalone ratio-test method.
- 19
- 18
- 18
- Creating value through IT-enabled integration in public organizations: A case study of a prefectural Chinese Center for Disease Control and Prevention18
The case reveals multiple types of IT values that surface in areas of managing immunization records, vaccine supply chain management, and public health crisis response, but the values only fully materialize after the research site underwent an evolutional path and a properly architected information system finally emerged to allow IT-enabled data and process integration.
- Character segmentation and retrieval for learning support system of Japanese historical books18
The experimental results show that the proposed character segmentation and retrieval method segmented all text lines correctly and successfully extracted more than 79% of the complicated characters, as well as provided promising character retrieval results.
- Joint Intrinsic and Extrinsic LiDAR-Camera Calibration in Targetless Environments Using Plane-Constrained Bundle Adjustment17
This paper introduces a novel targetless method for joint intrinsic and extrinsic calibration of LiDAR-camera systems using plane-constrained bundle adjustment (BA), which surpasses other state-of-the-art (SOTA) methods but also maintains remarkable calibration accuracy even within challenging environments.
- 16
- Moving Car Recognition and Removal for 3D Urban Modelling Using Oblique Images16
The proposed method is effective in recognizing and removing moving cars and can repair the geometric deformation and error texture mapping problems caused by moving cars.
- Structural Semantic Adversarial Active Learning for Image Captioning16
A structural semantic adversarial active learning (SSAAL) model that leverages both visual and textual information for deriving the most representative samples while maximizing the image captioning performance is proposed.
- 16
- 16
- Stochastic Context Consistency Reasoning for Domain Adaptive Object Detection15
A stochastic context consistency reasoning network with the self-training framework is proposed and an intra-changeable context consistency reasoning module is developed, which constructs an intra-context consistency paradigm to strengthen the utilization of context relations by utilizing pseudo labels to supervise the predictions of the student model.
- Adaptive Feature Selection With Reinforcement Learning for Skeleton-Based Action Recognition15
This article proposes a feature selection network (FSN) with actor-critic reinforcement learning that can cooperate with any existing GCN-based framework to enhance the recognition accuracy, the proposed GGCN can extract richer skeleton features for skeleton-based action recognition, and the method achieves superior performance over several public datasets.
- 14
- DFedSN: Decentralized federated learning based on heterogeneous data in social networks14
This paper sets up the affine distribution of user data structure to capture the users who have the heterogeneity between data, and puts forward an approach called DFedSN training to reduce the independent identically distributed data loss.
- 14
- Research on denoising method of chinese ancient character image based on chinese character writing standard model14
A Chinese ancient text image denoising method based on the Chinese character writing standard model that can increase the peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) of the image by at least 23.8% and 11.4%, and the user evaluation index has also reached more than 80%.
- 14
- End-to-end trainable network for superpixel and image segmentation14
An end-to-end trainable network is proposed that generates superpixels and segment images simultaneously and trains a differentiable clustering algorithm module to produce accurate superpixels.
- 14
- Super-resolution synthetic MRI using deep learning reconstruction for accurate diagnosis of knee osteoarthritis13
DLR synthetic knee MRI provided both weighted images for clinical diagnosis and accurate T2 measurements for more confidently identifying early cartilage degeneration from normal-appearing cartilages compared to the conventional morphologic knee sequences.
- Prosody-Enhanced Acoustic Pre-training and Acoustic-Disentangled Prosody Adapting for Movie Dubbing13
This work proposes an acoustic-prosody disentangled two-stage method to achieve high-quality dubbing generation with precise prosody alignment and incorporates an in-domain emotion analysis module to reduce the impact of visual domain shifts across different movies, thereby enhancing emotion-prosody alignment.
- Multi-Grained Representation Aggregating Transformer with Gating Cycle for Change Captioning13
The proposed MURAT can greatly improve the ability to describe the actual change in the distraction of irrelevant change and achieves state-of-the-art performance on three benchmarks, CLEVR-Change, CLEVR-DC, and Spot-the-Diff.
- An Improved Model on the Vague Sets-Based DPoS's Voting Phase in Blockchain13
This article improves the vague sets-based model of DPoS by introducing a new mapping from the vague set to the fuzzy set and assigning weights to each node, and considering that the influence factor of abstention vote increases with the increase of effective votes.
- Dynamic Contrastive Learning with Pseudo-samples Intervention for Weakly Supervised Joint Video MR and HD13
A proportional sampling strategy for pseudo-samples over random masks is introduced and dynamic intervention contrastive loss is designed to enhance the core feature-matching ability of the model dynamically.
- 13
- Joint image representation and classification in random semantic spaces13
A novel algorithm is proposed by combining image representation and classification in the random semantic spaces by combining local features with the sparse coding technique and using the encoding parameters for raw image representation to achieve better performances.
- 13
- 13
- 13
- Pick-and-Draw: Training-free Semantic Guidance for Text-to-Image Personalization12
Qualitative and quantitative evaluations demonstrate that the Pick-and-Draw approach significantly improves identity consistency and generative diversity, establishing a new Pareto frontier in the balance between subject fidelity and image-text alignment.
- 12
- 12
- Reverse Densely Connected Feature Pyramid Network for Object Detection12
The proposed Rev-Dense FPN is a novel multi-scale feature transformation and fusion method for object detection that directly fuse all the feature maps of higher levels than the current one and is more effective in dealing with diversified object sizes.
- 12
- Constructing a Collocation Learning System from the Wikipedia Corpus12
The design and development of a collocation learning system that is built from Wikipedia text and provides language learners with an easy-to-use interface for looking up collocations of any word that occurs in Wikipedia is described.
- 12
- 11
- 11
- 11
- Research on task offloading optimization strategies for vehicular networks based on game theory and deep reinforcement learning11
A task offloading decision mechanism (TODM) based on cooperative game and deep reinforcement learning (DRL) is proposed, which effectively reduces delay and energy consumption while ensuring task completion.
- 11
- Cross-media Topic Detection with Refined CNN based Image-Dominant Topic Model11
A novel image dominant topic model is proposed, which projects both the text modality and the visual modality into a semantic simplex and an improved CNN feature is introduced to capture more visual details by fusing the convolutional layer and fully-connected layer.
- Time evolving graphical password for securing mobile devices11
A new login option for unlocking mobile devices is proposed called Time-Evolving Graphical Password (TEGP), which improves the strength of the password gradually over time by evolving the distortion degree of the images in the challenge portfolio without changing the pass images.
- Person Re-Identification by Spatial Pyramid Color Representation and Local Region Matching11
This paper proposes a novel spatial pyramid color representation (SPCR) and a local region matching scheme, to explore person appearance for re-identification, and shows robustness to illumination changes and slight viewpoint variations.
- 11
- Vicept11
This paper proposes a new perspective called Vicept to associate elementary visual features and cognitive concepts and shows its effectiveness in concept detection task and can be extended to web-scale scenarios.
- 11
- Steganography: DCT Coefficients Reparation Technique in JPEG Image11
Semantic Tuplespaces (STS) is introduced based on a combination of Semantic Web technologies and tuplespace computing that acts as a middleware to enable communication and coordination based on the principle of publish and retrieve of semantic information.
- 11
- Learning Domain Invariant Features for Unsupervised Indoor Depth Estimation Adaptation10
This article proposes to study the generalization of depth estimation models across different indoor scenarios in an adversarial-based domain adaptation paradigm, and reconstructs depth maps from latent representations with the supervision of labeled source data.
- Automatic Relation-aware Graph Network Proliferation10
This work proposes Automatic Relation-aware Graph Network Proliferation (ARGNP) for efficiently searching GNNs with a relation-guided message passing mechanism and designs a network proliferation search paradigm to progressively determine the GNN architectures by iteratively performing network division and differentiation.
- A Structured Latent Variable Recurrent Network With Stochastic Attention For Generating Weibo Comments10
A structured latent variable recurrent network is proposed, which exploits the hierarchical-structured latent variables with stochastic attention to model the variations of comments and introduces a stochastic attention to infer the key-words of interest in the input post.
- Integrated optimization of urban agglomeration passenger transport hub location and network design10
The improved genetic algorithm is designed to solve the bi-level programming model and the model validity and algorithm feasibility is validated through the computing example.
- 10
- Sequential search based power allocation and beamforming design in overlay cognitive radio networks10
This work considers the problem of power allocation and linear beamforming design in overlay cognitive radio (CR) networks where the cognitive user is admitted to transmit simultaneously with the primary user and proposes algorithms based on uniform quantization and sequential search to solve the corresponding non-convex optimization problems.
- 10
- Aerial image model and application to aberration measurement10
A streamlined aerial image model that is linear with respect to projection optic's aberrations is presented, which allows for automatic identification of image's primary degrees of freedom, such as bananicity and Y-icity among others.
- FlowDubber: Movie Dubbing with LLM-based Semantic-aware Learning and Flow Matching based Voice Enhancing9
A novel dubbing architecture based on Large Language Model (LLM) and Conditional Flow Matching (CFM), named FlowDubber, which achieves high-quality audio-visual sync and pronunciation by incorporating a large speech language model with dual contrastive alignment while improving acoustic quality via Flow-based Voice Enhancing (FVE).
- 9
- 9
- 9
- 9
- A two-step approach to describing web topics via probable keywords and prototype images from background-removed similarities9
By removing background similarities, a devcovolution model is utilized to strip the similarities among non-informative words/images during web topic detection, and the background-removed similarities are reconstructed to identify the probable keywords and prototype images during topic description.
- LSH-based semantic dictionary learning for large scale image understanding9
An online semantic dictionary learning algorithm on the base of both locality sensitive hashing (LSH) and stochastic approximations, which can scale up to large scale datasets with millions of training samples and speedup the efficiency of follow-up processing is proposed.
- Palmprint Recognition Using Fourier-Mellin Transformation Based Registration Method9
A new method for palmprint recognition based on the registration results between two palmprint images is proposed, using the registration parameters obtained by the Fourier-Mellin Transformation to measure the similarity betweenTwo palmprint patterns.
- DFA-UNet: dual-stream feature-fusion attention U-Net for lymph node segmentation in lung cancer diagnosis8
Comprehensive experiments confirm the excellent segmentation effectiveness of the DFA-UNet combined attention mechanism for ultrasound images, underscoring its important significance for future research on medical images.
- A low-light-level image enhancement algorithm combining Retinex and Transformer8
The experimental results show that this low light image enhancement method based on Retinex theory and Transformer has achieved excellent performance in low light image enhancement tasks and has certain application value.
- Travel-Energy-Based Timetable Optimization in Urban Subway Systems8
A heuristic generic algorithm is developed to solve the optimization problem through a special binary coding method and the model is applied to the Yi-zhuang line in the Beijing subway system to obtain a passenger-oriented energy-minimizing timetable.
- 8
- 8
- 8
- Cross‐modal semantic correlation learning by Bi‐CNN network8
A novel cross modal retrieval framework, which integrates feature learning and latent space embedding, is proposed, which increases the discriminative ability among different categories, but also maximizes the relativity between different modalities.
- A Blockchain-Based Hierarchical Authentication Scheme for Multiserver Architecture8
The proposed blockchain-based authentication scheme provides a hierarchical authentication method to solve the problems of user permission control and user revocation caused by no registration center and is resistant to attacks such as impersonation attacks and man-in-the-middle attacks.
- 8
- Non-destructive identification of commercial jerky types based on multi-band hyperspectral imaging with machine learning7
The findings demonstrate short-wave-near-infrared hyperspectral imaging combined with linear models (logistic regression and Support Vector Machine with linear kernel parameter settings) is better for identifying the types of jerky.
- 7
- Diverter-Guider Recurrent Network for Diverse Poems Generation from Image7
An end-to-end probabilistic Diverter-Guider Recurrent Network (DG-Net), which is a context-based encoder-decoder generative model with the hierarchical stochastic variables, which can generate diverse poems with the poeticness and relevance.
- Structured Stochastic Recurrent Network for Linguistic Video Prediction7
This work introduces a new task of Linguistic Video Prediction (LVP), which aims to predict the forthcoming events based on past video content and generate corresponding linguistic descriptions, and proposes an end-to-end probabilistic approach named structured stochastic recurrent network (SRN) to characterize the one- to-many connections between past visual clues and possible future events.
- Character segmentation and transcription system for historical Japanese books with a self-proliferating character image database7
An interactive system for assisting transcription work for digitized historical woodblock-printed Japanese books published in the seventeenth to nineteenth centuries is described, and it was shown that when the number of character images in the database increased, the transcription accuracy also increased accordingly.
- 7
- Link reliability of underlay cognitive radio: symbol error rate analysis and optimal power allocation7
This study studies the SER performance in underlay cognitive radio systems comprising a secondary transmitter, a secondary receiver and a primary receiver and derives the optimal power allocation strategy to minimize the average SER at the SR.
- ID-based authentication scheme combined with identity-based encryption with fingerprint hashing7
A novel ID- based authentication scheme based on ID-based encryption (IBE) and fingerprint hashing method is proposed to solve the difficulties in the IBE scheme, which includes message receiver authenticating the sender, the trusted authority (TA) authenticates the users and transmitting the private key to them.
- 6
- MaTCR: Modality-Aligned Thought Chain Reasoning for Multimodal Task-Oriented Dialogue Generation6
A Modality-aligned Thought Chain Reasoning (MaTCR) framework to insert explicit reasoning process for multimodal task-oriented dialogue generation is proposed and extensive experiments over a benchmark dataset demonstrate that MaTCR outperforms the existing methods and provides stronger interpretability.
- 6
- 6
- 6
- Joint Detection of Tap and CEA Based on Deep Learning Medical Image Segmentation: Risk Prediction of Thyroid Cancer6
The experimental results show that the segmentation overlap rate of this method is as high as 98.4%, indicating that the algorithm proposed in this paper can more accurately extract the thyroid nodule area.
- Heuristic Depth Estimation with Progressive Depth Reconstruction and Confidence-Aware Loss6
A Heuristic Depth Estimation Network with progressive depth reconstruction and confidence-aware loss that leverages the reference samples with low confidence to distill the spatial geometric and local semantic information for dense depth prediction.
- Selective information passing for MR/CT image segmentation6
A novel 3D network with self-supervised function, named selective information passing network is proposed, which achieves improved segmentation results and outperformed other state-of-the-art methods.
- Enabling 5G: sentimental image dominant graph topic model for cross-modality topic detection6
A sentimental image dominant graph topic model is proposed, that can detect the topic from the heterogenous data and mine the sentiment of each topic and improve the discriminative power of CNN feature by jointly optimizing the output of both convolutional layer and fully-connected layer.
- Improving the Triple-Carrier Ambiguity Resolution with a New Ionosphere-Free and Variance-Restricted Method6
Experimental results indicate that the proposed method can obtain a better performance of ambiguity resolution, even though the double-differenced ionospheric delay increases up to 72.4 cm at the baseline of 68.8 km.
- Partitioned Sampling of Public Opinions Based on Their Social Dynamics6
This paper rigorously formulate the idea of partitioned sampling, which partitions individuals with high opinion similarities into groups and then samples every group separately to obtain an accurate estimate of the population opinion, and adapt an approximation algorithm and a heuristic algorithm to solve the optimization problem.
- 6
- 6
- 6
- 6
- 6
- Protocol for generating brain metastatic tumor cells through repeated intracardiac injections in mice5
A protocol for generating the lung cancer cell line, LLC1-BMT5, with highly brain metastatic tropism through multiple rounds of in vivo selection will facilitate the study of the molecular mechanisms of BrMs and the development of anti-BrM treatment strategies.
- Weighted progressive alignment for multi-source domain adaptation5
A novel framework of weighted progressive alignment (WPA) is proposed, in which a two-stage alignment with two distinct domain classifiers, as well as a dedicated classifier to judge the importance of source domain samples are developed.
- Research on Chinese ancient characters image recognition method based on adaptive receptive field5
An adaptive receptive field-based image recognition network for Chinese ancient characters that fully considers the high sensitivity of residual dense network to data fluctuation and uses selective convolution kernel to provide a dynamic selection mechanism for the receptive fields of different images, so that it can focus on the fine structure of Chinese characters and thus realize cross-channel information interaction and integration.
- Cloud computing based big data platform construction of university informatization teaching service in Yunnan Province under the background of community of destiny5
The big data platform model of university informatization intelligent teaching service of Yunnan University Teacher Education Alliance based on cloud computing is preliminarily constructed to provide reference for regional alliance education related research.
- Using normal dictionaries to extract multiple semantic relationships5
A rule-based semantic relationship extraction method for Chinese sentiment lexicon is proposed to solve the problem that words in the sentiment dictionary cannot be updated in time, and expand the lexicon and the semantic relationship in the dictionary.
- Local-binarized very deep residual network for visual categorization5
The results show that the proposed local binary residual block (LBB) model can accelerate the training and inference of the network with only a slight performance degradation.
- 5
- Transferrable Referring Expression Grounding with Concept Transfer and Context Inheritance5
A Concept-Context Disentangled network (CCD) is proposed which transfers concepts from auxiliary classification data with new categories meanwhile inherits context from REG data to ground new objects and outperforms the compared approach on the new-category test datasets.
- 5
- 5
- 5
- Image-regulated graph topic model for cross-media topic detection5
A novel Image-regulated Graph Topic Model (IGTM) is proposed, which combines cross-media data together in the modeling process and can achieve higher quality underlying topics as image relationships could serve as weakly-supervised information for topic modeling.
- Research and Design of Multi-Source Heterogeneous Information Integration Platform for Metadata Service5
This paper represented that the metadata service was the core of integration frame, and the realization of integration management platform was feasible.
- CDU linear model based on aerial image principal components5
This model expands linear representation to comprehensive image quality specs in a large focus-dose grid and identifies corrections to the full Bossung curve or process window shapes that are proportional to aberration levels.
- 5
- Collective Behavior Clone with Visual Attention via Neural Interaction Graph Prediction4
Experimental results show that this method outperforms previous approaches in predicting both the interaction graph and swarm actions with higher accuracy, and offers a promising approach for understanding interaction mechanisms and swarm dynamics in future swarm robotics research.
- CATS: A Pragmatic Chinese Answer-to-Sequence Dataset with Large Scale and High Quality4
A Unified Graph Transformation approach is proposed to establish a joint encoding space for the two hybrid knowledge resources and convert this task to a graph-to-text problem and demonstrate the effectiveness of this proposed method.
- Reducing Intrinsic and Extrinsic Data Biases for Moment Localization with Natural Language4
A hybrid multi-modal debiasing network with temporal consistency constraint for MLNL and the hybrid linguistic activating module with external knowledge to relieve the extrinsic bias is designed, which introduces a prior guidance to focus the discriminative information from the tail samples.
- 4
- Graph-to-Text Generation with Dynamic Structure Pruning4
A Structure-Aware Cross-Attention (SACA) mechanism to re-encode the input graph representation conditioning on the newly generated context at each decoding step in a structure aware manner is proposed.
- 4
- AGNNGRU-CDR: Attentive Graph Neural Networks and Gate Recurrent Unit for Cross-Domain Recommendation4
- 4
- 4
- Two-stream deep sparse network for accurate and efficient image restoration4
A two-stream sparse network to explicitly learn shallow and deep features to enforce their respective contribution to image restoration, and demonstrates that TSSN outperforms most of state-of-the-art methods on benchmark datasets on both quantitative metric and visual quality.
- 4
- Improved Video Moving Target Tracking Based on Camshift4
An improved algorithm based on Camshift algorithm, the similarity between target color and background color for the Cam shift algorithm, is proposed and the results verify the feasibility and effectiveness of the algorithm.
- The bearing of macro and micro culture on digital activity: EFL teachers' perception of computer-based matriculation oral language test reform4
The often-overlooked construct---culture at macro and micro levels in the testing context is discussed and the impact of cultural artifacts on the face-to-computer testing is modeled by connecting the ideas of culture to the framework of activity theory.
- Change detection based on similarity measurement of object histogram using high-resolution remote sensing imagery4
A new method introducing similarity measurement of object histogram is proposed and Experimental results show that the method can get results with higher accuracy in change detection, especially for high-resolution remote sensing images.
- 3d Geometric Model of Transparent Tower Equipment Display in OpenGL4
3d model of the tower equipment how to display dynamically in OpenGL and interface of OpenGL import 3d model is described in detailed, and format and reading file of STL (Stereo Lithography) are elaborated.
- 4
- 4
- Resolution of a Combinatorial Problem using Cultural Algorithms4
The authors show that Japanese Puzzles are constrained combinatorial optimization problems, which can be solved using Cultural Algorithms, and features, such the use of a belief space involve many proposed solutions and local search heuristics; can be taught using these puzzles.
- The application research of DPSA and large-scale system decomposition-coordination for cascaded hydroelectric short-term optimal scheduling4
Two methods, DPSA and Large-Scale System decomposition-coordination, are contrasted and analyzed in detail from calculating idea and process and both methods will arrive convergence because of the identical idea.
- 3
- Cognitive-inspired xLSTM for multi-agent information retrieval3
This work proposes a cognitive-inspired xLSTM model, specifically designed for multi-agent information retrieval, that significantly outperforms six state-of-the-art methods in terms of training time, inference time, and key performance metrics such as accuracy, recall, and F1 score.
- 3
- Exploring the Teaching Reform Path of Ideological and Political Education in Colleges and Universities in the Context of Deep Learning3
The positive impact of the bilinear blended teaching reform on enhancing student ideological engagement and promoting educational reform in colleges is affirmed, affirming the positive impact of the bilinear blended teaching reform.
- Domain Shared and Specific Prompt Learning for Incremental Monocular Depth Estimation3
This paper attributes the cross-domain catastrophic forgetting to the domain distribution shifts and continuous variations of depth space, and proposes Domain Shared and Specific Prompt Learning (DSSP) for incremental monocular depth estimation.
- An attention mechanism and GAN based low-light image enhancement method3
The experimental results show that the saliency model proposed in this paper can effectively enhance the low-light image and obtain a good visual effect.
- 3
- 3
- 3
- 3
- 3
- Vanilla Feedforward Neural Networks as a Discretization of Dynamical Systems3
This paper back to the classical network structure and proves that the vanilla feedforward networks could also be a numerical discretization of dynamic systems, where the width of the network is equal to the dimension of the input and output.
- 3
- 3
- 3
- 3
- 3
- 3
- A Personalized Recommendation Algorithm with User Trust in Social Network3
The user trust is integrated into the recommendation algorithm, by dividing the user trust into 2 parts: user score trust and user preference trust, which shows that the proposed algorithms can effectively improve the accuracy of recommendation.
- 3
- Template Tables and Computational Practices in Early Modern Chinese Calendric Astronomy3
A writing format, the ‘template table’ (suanshi, ) that was designed to guide the process of calendrical astronomical calculations in early modern China, became widespread from the Ming period (1368–1644) onwards.
- 3
- 3
- 3
- Digital image watermark algorithm based on compressive sensing3
A new algorithm of digital image watermark based on compressive sensing theory is proposed to embed watermark in the observation domain of the sparsed image and restore sparse signal.
- 3
- Primitive and Grain Estimation Using Flexible Magnification for a Morphological Texture Model3
Experimental results show that the proposed method provides more representative grains than the conventional method, and a suitable structuring element for the homothetic magnification process of the primitive is explored.
- 3
- 3
- A Usage-Model Based Approach to Test Therac-253
It is shown that the error-producing sequence in the Therac-25 can be generated through the proposed user model, which serves a base for acceptance testing.
- SafeCFG: Controlling Harmful Features with Dynamic Safe Guidance for Safe Generation2
This work proposes SafeCFG to adaptively control harmful features with dynamic safe guidance by modulating the CFG generation process, which can simultaneously modulate different harmful CFG generation processes, so it could eliminate harmful elements while preserving high-quality generation.
- 2
- From Language to Instance: Generative Visual Prompting for Zero-shot Camouflaged Object Detection2
LiP is the first framework that utilize text-to-image generative model to construct instance-level visual prompts in COD task, and is the first framework that utilize text-to-image generative model to construct instance-level visual prompts in COD task.
- 2
- 2
- The recovery strategy of interdependent networks under targeted attacks2
A novel recovery strategy that focuses on recovering a predetermined number of failed nodes that are adjacent to the largest connected component of each individual network, along with their dependent nodes, at each recovery stage, which significantly enhances the capacity to prevent system breakdowns for interdependent networks subjected to targeted attacks.
- Technique Report of CVPR 2024 PBDL Challenges2
This technical report summarizes the outcomes of the Physics-Based Vision Meets Deep Learning (PBDL) 2024 challenge, held in CVPR 2024 workshop, focusing on Low-Light Enhancement and Detection as well as High Dynamic Range (HDR) Imaging.
- 2
- 2
- 2
- 2
- 2
- 2
- Universal Semantic Tagging for English and Mandarin Chinese2
A set of language-specific semantic phenomena is discussed, a set of new annotation specifications are proposed, and a series of neural tagging models are evaluated to gauge how successful semantic tagging can be.
- 2
- SiamCenter2
A new object tracking model is proposed which combining Siamese network and anchor-free object detection algorithm, named SiamCenter, directly predicts center point location and size of object.
- 2
- Regularized topic-aware latent influence propagation in dynamic relational networks2
By integrating both topic and structure information into the regularized non-parametric probabilistic learning process, RegInfoIbp is more efficient and accurate in discovering the key factors in the content and influential users in dynamic network structure.
- A Longitudinal Study of Self-Regulated Learning in an Online Game-Based EFL Formative Assessment Blended Learning Environment2
The intent of the present research is to exemplify how SRL, an active area in educational psychology, can help to investigate the mechanism and process of SRL in the online formative assessment environment over a long period of time.
- 2
- 2
- 2
- The Design of a U-Turn Lane Right Set and Its Control Mode at an Intersection2
The implementation of u-turn lane right set and the improvement of control mode can effectively avoid conflict points between u- turn traffic and other traffic; delay time, stop time and queue length have been reduced significantly, and traffic flow capacity at intersection has also been greatly improved.
- 2
- Ontology Mapping Algorithm Based on the Regularization Framework of Graph2
The edge weights of the ontology mapping are computed with the kernel function and, using the regularization framework of graph, the optimized function is obtained and all the vertices of the Ontology graph are mapped as real numbers.
- 2
- Euclidean distance based method for-unclassifiable region of support vector machine2
The experimental results on benchmark datasets show that EDM eliminates the UR in conventional MSVM and improves the classification capacity and generalization ability of MSVM effectively.
- A Coarse-to-Fine Searching Method with Kernel Matching Based on Bhattacharyya Coefficients2
Aiming at object tracking in large motion area, a mean shift algorithm is proposed based on coarse-to-fine searching with kernel matching that can efficiently use a similarity measure function to realize the rough location of motion object.
- 1
- 1
- 1
- SSTrack: Sample-interval Scheduling for Lightweight Visual Object Tracking1
This work proposes SSTrack, a new easier-to-harder training schedule for the lightweight object tracker that achieves State-of-the-Art CPU-real-time accuracy on 5 UAV-tracking benchmarks and 5 general object tracking benchmarks.
- Frequency-aware Correlation Discovering and Spatial Forgery Clue Distilling for Synthetic Image Detection1
This paper designs frequency-aware correlation discovering module to learn dynamic filters by instance-adaptive frequency masking block for identifying prominent spatial deficiencies, which distributed in different spatial positions with various patterns.
- 1
- Research on the Resource Sharing Mode of Civic and Political Education Platform in Colleges and Universities in the Information Technology Era and the Path of Student Participation1
Correlation coefficients and regression models were used to explore the relationship between the resource sharing model of the Civic and Political Education Platform and the students’ participation path and revealed the relationship between students’ participation paths, which makes the ideological education in colleges and universities rise to a new level.
- 1
- 1
- Cross-modal representation learning and generation1
The character of multi - source and multi - modality of data has become a challenging problem in multimedia research and its representation and generation can be as two key factors in cross - modal learning research.
- Research on cross-modality person re-identification based on generative adversarial networks and modal compensation1
A cross-modal person re-identification method based on generative adversarial network and feature-level modal compensation is proposed to solve the problem of excessive image modal differences.
- 1
- 1
- Detection and Removal of Visible Watermarks Combined with Perceptual Attention1
Experimental results show that the proposed method of removing visible watermarks including detection and removal can improve the objective evaluation indexes such as peak signal to noise ratio (PSNR) and structural similarity.
- DSPNet: Towards Slimmable Pretrained Networks based on Discriminative Self-supervised Learning1
This paper extends the idea of slimmable networks to a discriminative SSL paradigm, by integrating SSL and knowledge distillation gracefully and shows comparable or improved performance of DSPNet on ImageNet to the networks individually pretrained one by one under the linear evaluation and semi-supervised evaluation protocols, while reducing large training cost.
- 1
- 1
- 1
- Multi-level Data Fusion Algorithm towards Privacy Protection in Wireless Sensor Networks1
Results show that the mechanism proposed in this paper can significantly reduce resource overheads, effectively guarantee data security and fusion efficiency.
- Secure Data Aggregation Mechanism based on Constrained Supervision for Wireless Sensor Network1
Results show that the proposed mechanism can effectively identify the unreliable nodes, guarantee the system security and prolong the network lifetime.
- Development of Post-production Synthesis Technology1
The traditional and current post-production industry is analyzed, and the current development process through the prospect of future film and television synthesis technology is proposed.
- 1
- 1
- Integration of QR Code Based on Generation, Parsing and Business Processing Mechanism1
The integration of QR code technology has become a key breakthrough point to improve the competitiveness of enterprises and the process of information and transformation of society.
- 1
- 1
- 1
- 1
- 1
- Research of Computer Assisted English Teaching Based on Corpus1
Corpus linguistics is not a new subject, but a new study method, with real language data as the research object, analyzing a large number of language facts from the macroscopic Angle, and finding the rule of language in it.
- Energy-efficient Buffer-aided Optimal Relay Selection Scheme with Power Adaptation and Inter-relay Interference Cancellation1
Numerical results show that the proposed scheme outperforms other relay selection schemes in terms of outage behavior with power adaptation and IRI cancellation in the same relay number and buffer size scenario.
- 1
- Efficient virtual network transmission using correlated equilibrium on Xen-based platform1
This paper proposes an approach of redesigning the upgrade mechanism, and further rebuilding the structure of the patch file to solve the virtual network transmission problems, and make redesigned upgrade mechanism more efficient than the current one.
- 1
- Pairwise Co-betweenness for Several Types of Network1
This paper analyzes the pairwise co-betweenness of WS network model with the different reconnection probability which including rule, smallworld and random network, and obtains vertex-induced subgraph with the highest betweenness vertices, and the edge- induced subgraphwith the highest pairwiseCo- Betweenness edges.
- Design of a Lightweight Automatic Question Answering System in Education Information Service Domain Based On FAQ1
The paper introduces the design and implementation of a lightweight automatic question answering system in education information service domain based on FAQ (Frequently Asked Question), and the workflow and structure of the system were described, and the key techniques were expounded.
- Polysemious visual representation based on feature aggregation for large scale image applications1
Through the proposed framework of generating polysemious image representation through three levels of feature aggregation, multiple feature fusion and multiple semantic descriptions are both achieved in an integrated way.
- Fusing multi-cues description for partial-duplicate image retrieval1
This paper proposes a rotation-invariant Local Self-Similarity Descriptor (LSSD), which captures the internal structural layouts in the local textural self-similar regions around interest points, and formulate the Semi-Relative Entropy as the distance metric.
- An Immersive Environment for a Virtual Cultural Festival1
The proposed system combines vision, sound, immersive display, and real time interaction, which enables the users to feel as if they are actually participating in the parade.
- 1
- 1
- 1
- 1
- 1
- 1
- Pornographic image detection in multimedia messaging service1
A pornographic information filtering system for MMS that employs the area of skin region, the ratio of face in body, focus of photography, and the identification results of sensitive organs to measure the sensitivity of the images is presented.
- 1
- 1
- Fast outliers mining algorithm based on unit cell1
Experimental results show that the fast outliers mining algorithm proposed by partitioning the data into a set of units cell firstly is fast and effective and can deal with the large scalability of points.
- 1
- Comments and Corrections1
It is shown that the correct conclusion about phase noise is the original one in a frequency-domain analysis derived through a frequency-domain analysis.
- Bench Blasting Designed by Fuzzy Neural Network1
The application of fuzzy neural network model to intelligent design of bench blasting combines the advantages of computer with that of expertise to gain the experience/knowledge from what the experts did in actual blasting process in the past.
- 1
- General Model with Parameters Analysis of Syntax Tagging1
A general model of syntax tagging is presented, and the parameters of the syntax tagging models based on PSG and DG arezes, to try to improve on thentax tagging model based on DG.
- 1
- A Novel Identity Authentication Technique Without Trustworthy Third-Party Based on Fingerprint Verification1
This paper focuses on identity authentication techniques that authenticate users' identity relying on tokens or keys and one or many trustworthy third-party(s) that require databases running online, with multiple points of vulnerability and low efficiency.
- A recurrence algorithm of FFT1
This paper improves FFT algorithm into a recurrence expression, a (j) k=a (j-1) x+a ( j- 1) yω p, in transforming, where x,y and p can calculate with binary bit.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Reflections Based on the Latest 3D Films: The Development and Challenges of the Contemporary Chinese Animation Industry–
The paper analyzes from the perspectives of technology, art and industry, and points out that the application of 3D technology is an inevitable trend in the development of China's animation industry.
- –
- Protective Self-Adaptive Pruning to Better Compress DNNs–
This work proposes a Protective Self-Adaptive Pruning (PSAP) method, which merely depends on weights and gradients of model itself, instead of requiring an additional monitor as in early works.
- –
- –
- –
- A STUDY ON PHYSIOLOGICAL AND EMOTIONAL BEHAVIOR CHANGES OF EMOTIONAL INTERACTION IN SERIOUS GAMES–
The research shows that physiological signals can be used as an index of players' psychological state to a certain extent and comes to the conclusion that academic performance has little impact on the evaluation of learning.
- –
- –
- –
- Image Risk Assessment of the Thyroid Cancer Model Based on Discriminant Analysis and the Value of TAP and CEA Combined Detection–
The results showed that the levels of serum TAP and CEA in patients with thyroid cancer were significantly higher than those in Patients with benign thyroid diseases, and the difference was statistically significant (P < 0.05).
- –
- –
- Prediction of Function Removal Propagation in Linux Evolution–
The life cycle feature and the prediction heuristics presented in this paper can be utilized to facilitate the maintenance of large-scale complex software systems.
- –
- Active Perception Network for Salient Object Detection–
Active Perception Network (APN) is proposed to enhance inter-feature consistency for salient object detection and aggregate global information, which brings strong semantic consistency to the whole network.
- –
- Research on Fire Dispersion Data Processing Method Based on Cloud Model–
The method of cloud model is introduced to evaluate the data of the dispersion of one certain type of ammunition, and generates a comprehensive cloud model according to the established evaluation criteria.
- Analysis of the Spread of Experimental Animated Short Films in the Context of New Media–
With the birth of new media, including computer media, online media, virtual reality media, and augmented reality, the spread efficiency and channels of communication have been expanded and the content and quality of the experimental animated short films should be improved.
- Logical Parsing from Natural Language Based on a Neural Translation Model–
This paper proposes an approach to learn from denotations based on the Seq2Seq model augmented with attention mechanism that can successfully infer the correct logical forms and learn a meaningful semantic parser.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Information Cascades on Arbitrary Topologies–
The result shows that while both too little and too much information sharing causes nodes to take wrong decisions, for exactly the right amount of information sharing, asymptotically everyone can be right.
- –
- –
- A run-time optimization approach for reducing data movements using locality-aware searching–
Rules showing that the run-time data reuse optimization can effectively relieve the bottleneck are obtained and a run- time optimization framework for data reuse, called R-Tracker is proposed, which uses locality-aware searching approach to handle reuses.
- –
- –
- A Novel Approach Based on Multi-Layer Background Subtraction and Seed Region Growing for Sport Graphics Segmentation–
A robust multi-layer background subtraction technique and seed region growing approach which takes advantages of local texture features represented by local binary patterns (LBP) and photometric invariant color measurements in RGB color space to provide fast segmentation with high perceptual segmentation quality.
- –
- Camera Calibration Method Based on Halcon–
A nonlinear image geometrical model for array CCD camera was build and the interior parameters and exterior parameters as well were analyzed and this method is simple and it improves the calibration precision and computation speed, and has a good cross-platform portability.
- The Progression of the Visual Communication Design in the New Period–
This paper first studies the development of visual communication design process, then analysis under the background of new era should be how to innovate the visual Communication design, so it has certain practical significance and guiding significance.
- Impact Analysis of Information Teaching on Physical Education Based on Cluster Analysis–
The researcher analyzed the effectiveness of the physical education by combining with the influence factors of the effectiveness and by means of cluster analysis to proposed significant basis for physical education in how to teach students in accordance of their aptitude and how to improve the effectiveness.
- –
- Implementations and Mechanism of Dynamic Infrared Pattern Painting Camouflage–
The results show that dynamic IR pattern painting camouflage can effectively fails IR correlation tracking and it will be a developing direction.
- OIPBP: On-Demand Interdomain Path Building Protocol–
This paper extends BGP by adding more policies options into routing advertisements, and the inserted policies can be referred by the intermediate nodes when selecting paths.
- "Southern Elite" Game Event Produce Based On RPG Maker XP–
The use of software to make RMXP RPG game is described, through the production of events to achieve the game's storyline.
- –
- The Design Scheme of Public Service Platform Based on SaaS–
The article presents the design scheme of constructing public service platform which integrates some software application services based SaaS, and elaborate the key technologies including interface interaction, asynchronous call, and multi-tenant design and security strategy in build process of platform.
- –
- The Lightweight Process Improvement Framework Research Based on Knowledge Management–
The lightweight process improvement framework based on knowledge management is proposed after analyzing heavyweight process, agile methodologies and the knowledge management and experiments show that the framework can satisfy the software process improvement from the middle and small scale software companies.
- Attack Graph Optimization Based on Attack Distance–
The experimental results show that the optimized attack graph retains the most likely attack path and simplifies the attack graph.
- –
- –
- Video motion estimation with temporal coherence–
This paper formalizes the estimation of flow fields as an energy minimization problem and adapt the iterative reweighted least square (IRLS) to solve it and considers the Charbonnier penalty to make the penalties robust to outliers.
- –
- Speech Endpoint Detection Algorithm Based on First-order Finite Differential Quotient–
A method based on first-order finite differential quotient with a binary search to determine the endpoint detection improves the accuracy of endpoint detection and can achieve better voice activity detection in a noisy environment.
- –
- The Analysis and Comparison of SSL and SET in Electronic Commerce–
The paper analyzed the theory and the business process on the two security protocols of electronic payment: SSL and SET and gave the future development of SET protocol and the SET protocol cannot be applied widely in recent years.
- –
- On How to Deeply Seek Out User's Demand Information in Library-provided Personalized Services–
It is suggested that date base statistics, web data obtainment, feedback and exchange visits be employed to obtain user's demand information and a comparison between these three methods be made in order to provide a consult for libraries to personalize their services and build a comprehensive and effective users' demand information obtainment system.
- –
- Deep Research and Practice on Bilingual Education of Computer Science Brief Introduction–
The paper mainly introduces bilingual teaching mode and method of computer science brief introduction as well as some issues to which should be paid more attention.
- Study on Risk Measurement of Large-scale Project Based on Risk Network–
A risk measurement method that makes a clear and reasonable combination of the amount of risk loss and the loss probability and it can be used to measure the objective risk as well as to measure subjective risk is brought forward.
- Geographic routing algorithm based on virtual coordinate system–
DGA was applicable for more accurate network models, such as the network model based on the log-normal shadowing model, which is substantially different to the GPSR algorithm.
- Synthesis and Photophysics of Two Dipolar Copolymers Containing Hole-Transporting Unit and Alq3 Unit–
- –
- –
- Empirical study of a mathematical module teaching–
The background of mathematics modules with my math module teaching hospital status, and in my home computer professional as an example, select one of the 40 samples of a VB programming module with the three math to do a specialized component analysis of empirical research.
- A New Method for Camera Composition–
A constraint-based approach to camera composition is proposed, and quantitative criteria are given to evaluate the performance of the method, showing that images generated by the method can satisfy the basic requirement in animation generation.
- Discussion of the Public Security Informationization–
The modern public security system must establish an improved information system in order to respond to the social security issues, which can significantly promote the working efficiency and strength of the police.
- An Study on Several Issues of Detection Guided by Information–
How to change the present negative situation of detection work, convert passive fighting into initiative attack, and introduce the mechanism of information guiding detection into detection work to enhance the efficiency and accuracy of Detection work has become the consensus of many experts and scholars.
- Realizing initial evaluation of sci-tech journals by manifold approaches–
The authors analyze the issues of manuscript initial evaluation and put forward four approaches, for the purpose of standardizing initial evaluation of sci-tech journals.
- –
- –
- Schedule arrangement based on simulated annealing algorithm–
The simulated annealing algorithm applies to the schedule arrangement that requires symmetrical and the control parameter in the algorithm can quicken the convergence rate to obtain the optimum solution or approximate optimum solution.
- –
- Multi-scale Palmprint Recognition Using Registration Information and 2D Gabor Feature–
A novel method for palmprint recognition based on registration information and 2D Gabor features is described, which is used to guide ROI extraction and demonstrates the effectiveness of the method.
- The study on genetic algorithm on mining quantitative association rules–
A genetic algorithm is proposed in the present paper to solve the clustering problem which can be solved by FCM (Fuzzy Clustering Method), so as to avoid the local optimization that often occurs in FCM.
- An Improved Algorithm for Ellipses Detection Based on Randomizea Hough Transform–
A randomized Hough transform algorithm for ellipses detection (RED) is presented, which firstly selects three edge pixels randomly in the image, and takes them as the centers of three windows with the same size.
- Improving Fingerprint Recognition Performance Based on Feature Fusion and Adaptive Registration Pattern–
This paper combines minutiae, associate ridges and orientation fields to determine the registration pattern between two fingerprints and match them and demonstrates that the proposed algorithm is an accurate one.
- A navigaton and interactive design of Electronic Simulation Technique Network Course–
Taking Electronic Simulation Technique as an example, the navigation interface design of Electronic Simulation technique Network Course is introduced and the mutual interface design and the interactive mode are presented.
- –
- Research on Object- Oriented Continuous System Simulation–
The object- oriented analysis model of the continuous system simulation is developed by using the Coad/Yourdon method, and the methods of design and implementation are presented.
- How to Build Multidimensional Data Model Basing on Relational DB–
The dimension table of star mode is decompose into hierarchical, normalized and small dimension tables so that it would be easy to drill down and roll up which is generally called data mining.
- –
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-14. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.