Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks243 from public data
- A reinforcement learning approach to job-shop scheduling480
Reinforcement learning methods are applied to learn domain-specific heuristics for job shop scheduling to suggest that reinforcement learning can provide a new method for constructing high-performance scheduling systems.
- Reconstruction Network for Video Captioning358
A reconstruction network with a novel encoder-decoder-reconstructor architecture, which leverages both the forward (video to sentence) and backward (sentence to video) flows for video captioning, and can boost the encoding models and leads to significant gains in video caption accuracy.
- PUGAN: Physical Model-Guided Underwater Image Enhancement Using GAN With Dual-Discriminators351
This paper proposes a physical model-guided GAN model for UIE, referred to as PUGAN, and designs the Dual-Discriminators for the style-content adversarial constraint, promoting the authenticity and visual aesthetics of the results.
- UAV-Human: A Large Benchmark for Human Behavior Understanding with Unmanned Aerial Vehicles282
A fisheye-based action recognition method that mitigates the distortions in fisHEye videos via learning unbounded transformations guided by flat RGB videos is proposed, and experiments show the efficacy of this method on the UAV-Human dataset.
- 216
- Controllable Video Captioning With POS Sequence Guidance Based on Gated Fusion Network190
A gating strategy is proposed to dynamically and adaptively incorporate the global syntactic POS information into the decoder for generating each word, which not only boosts the video captioning performance but also improves the diversity of the generated captions.
- Computerized detection of clustered microcalcifications in digital mammograms using a shift‐invariant artificial neural network184
This study applies a shift-invariant neural network to eliminate false-positive detections reported by the CAD scheme, considerably better than that obtained in the previous study using a conventional three-layer, feed-forward neural network.
- 164
- Frequency Perception Network for Camouflaged Object Detection158
This work proposes a novel learnable and separable frequency perception mechanism driven by the semantic hierarchy in the frequency domain that achieves competitive performance in three popular benchmark datasets both qualitatively and quantitatively.
- High-Performance Job-Shop Scheduling With A Time-Delay TD(λ) Network133
Experimental tests show that this TDNN-TD(λ) network can match the performance of the previous hand-engineered system, and both neural network approaches significantly outperform the best previous (non-learning) solution to this problem.
- An improved shift‐invariant artificial neural network for computerized detection of clustered microcalcifications in digital mammograms130
Modifications were made to improve the performance of the SIANN and the zero-mean-weight constraint and training-free-zone techniques have been developed and a cross-validation training method was also applied to avoid the overtraining problem.
- Head Fusion: Improving the Accuracy and Robustness of Speech Emotion Recognition on the IEMOCAP and RAVDESS Dataset125
This work implemented an attention-based convolutional neural network (ACNN) model and conducted experiments on the Interactive Emotional Dyadic Motion Capture (IEMOCAP) data set and proposed a method called Head Fusion based on the multi-head attention mechanism to improve the accuracy of SER.
- Point-aware Interaction and CNN-induced Refinement Network for RGB-D Salient Object Detection99
This work introduces CNNs-assisted Transformer architecture and proposes a novel RGB-D SOD network with Point-aware Interaction and CNN-induced Refinement (PICR-Net), which achieves competitive results in both quantitative and qualitative comparisons.
- 91
- 88
- 87
- Optimally weighted wavelet transform based on supervised training for detection of microcalcifications in digital mammograms78
A technique for optimizing the weights at individual scales in the wavelet transform to improve the performance of the CAD scheme based on the supervised learning method is developed.
- 73
- 62
- 57
- Long-range terrain perception using convolutional neural networks51
A deep convolutional neural network model is integrated with a near-to-far learning strategy to improve the accuracy of terrain segmentation and make it more robust against wild environments.
- Reduction of false positives in computerized detection of lung nodules in chest radiographs using artificial neural networks, discriminant analysis, and a rule-based scheme51
A computer-aided diagnosis (CAD) scheme is being developed to identify image regions considered suspicious for lung nodules in chest radiographs to assist radiologists in making correct diagnoses using an artificial neural network, discriminant analysis, and a rule-based scheme.
- CM-YOLO: Typical Object Detection Method in Remote Sensing Cloud and Mist Scene Images50
A novel target detection method named CM-YOLO that integrates background suppression and semantic context mining, which can achieve accurate detection of targets under different cloud and mist conditions and achieves target feature enhancement.
- Query-Guided Prototype Evolution Network for Few-Shot Segmentation50
The Query-guided Prototype Evolution Network (QPENet) is presented, a new method that integrates query features into the generation process of foreground and background prototypes, thereby yielding customized prototypes attuned to specific queries.
- 49
- 49
- 48
- 48
- 40
- Image processing of human corneal endothelium based on a learning network38
A neural network was trained to extract the cell's boundary by showing part of the photomicrograph and its subjective boundary image, which is an outline drawing made by hand, and showed good performance with the microphotograph that was not trained.
- Residual Distillation: Towards Portable Deep Neural Networks without Shortcuts37
This work proposes a novel joint-training framework to train plain CNN by leveraging the gradients of the ResNet counterpart, and demonstrates that the plainCNN network without shortcuts can achieve the same level of accuracy as that of the ResNet baseline while achieving about 1 .
- Explainable zero-shot learning via attentive graph convolutional network and knowledge graphs36
This paper focuses on explainable ZSL, and presents a knowledge graph (KG) based framework that can explain the transferability of features in ZSL in a human understandable manner, and generates promising explanations for the transferable of features, and at the same time improves the recognition accuracy.
- DeepPrep: an accelerated, scalable and robust pipeline for neuroimaging preprocessing empowered by deep learning34
Evaluated on over 55,000 scans, DeepPrep demonstrates a 11-fold acceleration, exceptional scalability, and robustness compared to the current state-of-the-art pipeline, providing a promising solution to meet the scalability requirements of neuroimaging.
- 34
- 34
- 34
- Reading Relevant Feature from Global Representation Memory for Visual Object Tracking33
A novel tracking paradigm, consisting of a relevance attention mechanism and a global representation memory, which can adaptively assist the search region in selecting the most relevant historical information from reference features, is proposed.
- Learning Common Semantics via Optimal Transport for Contrastive Multi-View Clustering32
This paper presents CSOT, namely Common Semantics via Optimal Transport, to boost contrastive multi-view clustering via semantic learning in a common space that integrates all views and proposes a semantic-aware re-weighting strategy to treat samples differently according to their semantic significance, which improves the effectiveness of cross-view contrastive representation learning.
- 32
- Experience Improves Feature Extraction inDrosophila32
The results indicate that previous experience can enhance visual feature extraction in Drosophila and that MBs are required for this experience-dependent visual cognition.
- Learning Robust Representations with Graph Denoising Policy Network30
Experimental results on node classification task demonstrate the effectiveness of GDNet, outperforming the state-of-the-art graph representation learning methods on several well-studied datasets.
- Attention to Head Locations for Crowd Counting29
A novel method using an attention model to exploit head locations which are the most important cue for crowd counting, and introduces a relative deviation loss to compensate a commonly used training loss, Euclidean distance, to improve the accuracy of sparse crowd density estimation.
- Globally consistent alignment for planar mosaicking via topology analysis29
A generic framework for globally consistent alignment of images captured from approximately planar scenes via topology analysis, capable of resisting the perspective distortion meanwhile preserving the local alignment accuracy is proposed.
- 28
- Learning Actions from Human Demonstration Video for Robotic Manipulation28
Experimental results show that the proposed novel deep model to learn actions from human demonstration video for robotic manipulation could produce more accurate command from video demonstration than state-of-the-art work, thereby leading to more robust grasping performance.
- 27
- 26
- SDDNet: Style-guided Dual-layer Disentanglement Network for Shadow Detection26
A Style-guided Dual-layer Disentanglement Network (SDDNet) to model these layers independently, and devise a Feature Separation and Recombination (FSR) module that decomposes multi-level features into shadow-related and background-related components.
- 26
- 26
- PolypMixNet: Enhancing semi-supervised polyp segmentation with polyp-aware augmentation25
The proposed PolypMixNet, a semi-supervised framework for colorectal polyp segmentation, utilizing novel augmentation techniques and a Mean Teacher architecture to improve model performance, introduces the polyp-aware mixup (PolypMix) algorithm and incorporates dual-level consistency regularization.
- Rapid measurement of a high step microstructure with 90° steep sidewall25
Based on the new STM system, sample-tilt-scanning methodology is proposed for eliminating the scanning blind region between the probe and the microstructure, and it provides an effective and nondestructive solution for the measurement of high step or deep trench microstructures.
- 24
- Down to the Last Detail24
A Tree-Block (tree dilated fusion block) to replace standard ResNet-block to address the long-range translation and rich-details generation, and a multi-stage framework to synthesize person images, where fine-grained details can be well preserved.
- 23
- Hierarchical Perception-Improving for Decentralized Multi-Robot Motion Planning in Complex Scenarios22
- Snap-and-ask22
This paper addresses the problem of visual-to-text naming through the paradigm of answering-by-search in a two-stage computational framework, which is composed out of instance search (IS) and similar question ranking (QR).
- Circular Accessible Depth: A Robust Traversability Representation for UGV Navigation21
This article presents the circular accessible depth (CAD), a robust traversability representation for an unmanned ground vehicle (UGV) to learn traversability in various scenarios containing irregular obstacles, and proposes a neural network, namely CADNet with an attention-based multiframe point cloud fusion module, stability-attention module (SAM), to encode the spatial features from point clouds captured by LiDAR.
- 21
- Omni-word Feature and Soft Constraint for Chinese Relation Extraction21
An Omniword feature and a soft constraint method for Chinese relation extraction that makes a better use of sentence information and minimize the influences caused by Chinese word segmentation and parsing are proposed.
- Data-oriented locality sensitive hashing21
Data-Oriented LSH is proposed to reduce memory consumption when data are non-uniformly distributed and focused on the hash table construction, and thus the query-directed methods can be applied to the index to improve further.
- 20
- 20
- Exploring the Task Cooperation in Multi-goal Visual Navigation20
This work introduces two new designs to the reinforcement learning scheme: inverse dynamics model (InvDM) and multi-goal co-learning (MgCl), proposed to capture the navigation-relevant association between state and goal, and provide additional training signals to relieve the sparse reward issue.
- 20
- MMReason: An Open-Ended Multi-Modal Multi-Step Reasoning Benchmark for MLLMs Toward AGI19
This work introduces MMReason, a new benchmark designed to precisely and comprehensively evaluate MLLM long-chain reasoning capability with diverse, open-ended, challenging questions and designs a reference-based ternary scoring mechanism to reliably assess intermediate reasoning steps.
- Source localization and functional network analysis in emotion cognitive reappraisal with EEG-fMRI integration19
The proposed multimodal integration method for source localization identified the key cortices involved in emotion regulation, and the network analysis demonstrated the important brain regions involved in the cognitive control of reappraisal.
- Electrochemical Corrosion Behavior and Mechanical Properties of Nanocrystalline Ti–6Al–4V Alloy Induced by Sliding Friction Treatment18
The NG Ti–6Al–4V sample exhibited the high yield strength (956 MPa) with a moderate elongation of 8%.
- 17
- A tightly-coupled LIDAR-IMU SLAM method for quadruped robots16
A tightly coupled SLAM algorithm that integrates LIDAR and IMU sensors is proposed that significantly outperforms the LOAM algorithm, the NDT-based SLAM algorithm and the LeGO-LOAM algorithm in terms of positioning accuracy and exhibits superior robustness.
- 16
- Analysis of the inter-dataset representation ability of deep features for high spatial resolution remote sensing image scene classification16
An intensive investigation in the feature representation ability of such deep features from the aspect of inter-dataset scene classification of remote sensing images and conclusions are drawn regarding the strength and weakness of the CNN features in the application ofRemote sensing image scene classification.
- 16
- Review of cervical cell segmentation15
This review summarizes and analyzes the recent research on cervical cell segmentation, and includes an introduction to cervical cell segmentation datasets, commonly used evaluation metrics, and various segmentation methods.
- 15
- 15
- Learn by Observation: Imitation Learning for Drone Patrolling from Videos of A Human Navigator15
An imitation learning method for autonomous drone patrolling based only on raw videos that allows the drone to fly at a high altitude with a broad view and low risk, and carry out the integration of available user instructions and autonomous patrolling control commands.
- 14
- URSA: Precise Capacity Planning and Fair Scheduling based on Low-level Statistics for Public Clouds14
URSA, a capacity planning and fair scheduling system that is comprised of an online capacity planner, a performance interference estimator, and a contention-aware scheduling engine to eliminate unfair performance interference between the co-located workloads.
- Underwater Image Restoration by Means of Blind Deconvolution Approach14
This work provides an image enhancement method by blind deconvolution method based on small angle approximation for underwater laser imaging system by intentionally changing the MTF parameters and using non-model-based PSF as the initial guess.
- Nowhere to Go: Benchmarking Multirobot Collaboration in Target Trapping Environment13
A new benchmark to evaluate multirobot collaboration in target trapping environment (T2E) is proposed, multiple learning-based baselines in T2E are presented, and insights into regimes of multirobot collaboration are provided.
- 13
- 13
- 13
- 12
- SCE-SLAM: a real-time semantic RGBD SLAM system in dynamic scenes based on spatial coordinate error12
SCE-SLAM is presented: a novel real-time semantic RGB-D SLAM system that is built on theRGB-D mode of ORB- SLAM3 and significantly improves the localization accuracy and system robustness in all kinds of dynamic environments.
- 12
- 11
- 11
- 11
- 11
- 11
- Binary matrix factorization via collaborative neurodynamic optimization11
This paper presents a collaborative neurodynamic optimization approach to binary matrix factorization based on the original combinatorial optimization problem formulation and quadratic unconstrained binary optimization problem reformulations.
- 11
- 11
- Human-Guided Data Augmentation via Diffusion Model for Surface Defect Recognition Under Limited Data10
- 10
- Exploiting Inter-Sample Affinity for Knowability-Aware Universal Domain Adaptation10
A knowability-based labeling scheme which can be divided into two steps: knowability-guided detection of known and unknown samples based on the intrinsic structure of the neighborhoods of samples, and label refinement based on neighborhood consistency to relabel the target samples.
- SAR image super-resolution using deep residual SqueezeNet10
Experimental results demonstrate that the new algorithm brought forward in this paper not only has lightweight construction, but also can be faster, and get similar or better results than state-of-the-art methods.
- 10
- 9
- 9
- 9
- 9
- 9
- Three Dimensional AUV Complete Coverage Path Planning with Glasius Bio-inspired Neural Network9
Through simulation experiment and discussion, it can be verified that, whether it is a static environment or the dynamic environment for the whole three-dimensional space, AUV can cover all the specified work waters without omissions and collision.
- Autonomous and Adaptive Role Selection for Multi-Robot Collaborative Area Search Based on Deep Reinforcement Learning8
A hierarchical framework designed for the simultaneous mapping and searching of targets in multi-robot area search tasks and employs adaptive role learning through deep reinforcement learning, while the low-level primitive policy focuses on planning actions based on the designated roles.
- 8
- 8
- 8
- 8
- 8
- Research on camouflage target detection method based on improved YOLOv58
The improved CM-YOLOv5s algorithm has better detection effect, and realizes accurate identification and rapid positioning of military camouflage targets in complex environments.
- 8
- 8
- 8
- 8
- Microbial remediation as a key pathway for pollutant removal in constructed wetlands: Mechanisms, bioaugmentation strategies, and perspectives7
This work synthesizes microbial remediation mechanisms, from key genes and enzyme to catabolic pathways, for both conventional and emerging pollutants, and proposes advanced strategies including synthetic microbial consortia, precision biostimulants, and 3D-printing microbial immobilization for next generation wetland bioremediation systems.
- Contrastive Learning with Negative Sampling Correction7
Positive-Unlabeled Contrastive Learning treats the generated negative samples as unlabeled samples and uses information from positive samples to correct bias in contrastive loss, proving that the corrected loss used in PUCL only incurs a negligible bias compared to the unbiased contrastive loss.
- 7
- 7
- Intrusive Detection Systems Design based on BP Neural Network7
An intrusion detection system was constructed on the basis of the characteristics of BP neural network model and this eigenvector is presented to the neural network classification engine, as the input vector of a neural network.
- 6
- 6
- Recognition of a Cracked Hen Egg Image Using a Sequenced Wave Signal Extraction and Identification Algorithm6
In order to develop a new recognition method for cracked hen egg with high accuracy and adoption, images with different crack width of the hen eggs were captured and a sequenced wave signal extraction and identification algorithm was developed.
- 6
- 6
- 5
- Learning Instruction-Guided Manipulation Affordance via Large Models for Embodied Robotic Tasks*5
This paper proposes Instruction-Guided Affordance Net (IGANet) for predicting affordance maps of instruction-guided robotic manipulation tasks by utilizing powerful priors from vision and language encoders pre-trained on large-scale datasets and develops a Vison-Language-Models-based data augmentation pipeline, which can generate a large amount of data automatically for model training.
- 5
- 5
- Beyond Sum and Weighted Aggregation: An Efficient Mixed Aggregation Method with Multiple Weights for Image Search5
This paper presents a general mixed aggregation method that unifies sum and weighted aggregation methods, and proposes to compute multiple weighting coefficients rather than one for each to be aggregated vector by partitioning it into several components.
- 5
- Asymmetric Information Enhanced Mapping Framework for Multirobot Exploration Based on Deep Reinforcement Learning4
This article proposes AIM-Mapping, an Asymmetric InforMation enhanced Mapping framework based on deep reinforcement learning that fully leverages the privileged information to help construct the environmental representation as well as the supervised signal in an asymmetric actor–critic training framework.
- 4
- 4
- Ship Detection via Dilated Rate Search and Attention-Guided Feature Representation4
A dilated convolution parameter search strategy is presented to adaptively select the dilated rate for the multi-branch extraction architecture, adaptively obtaining context information of different receptive fields without sacrificing the image resolution.
- RETRACTED ARTICLE: Simulation of cross-modal image-text retrieval algorithm under convolutional neural network structure and hash method4
The results reveal that the loss function of adaptive weight based on hash network reduces the influence of imbalances positive and negative sample on retrieval performance, the constrained regular terms reduce the error caused by quantization, avoids information loss caused by "Relaxation" strategy in traditional methods, and optimizes hash structure.
- 4
- 4
- 4
- 4
- Ant colony algorithm based scheduling for handling software project delay4
An ant colony algorithm is proposed to address the problem of employee scheduling which shows that in the scheduling process, one activity can have many ways for selecting employees from another project and it is impractical to generate a search tree for implementation.
- 4
- 4
- 3
- HRL-KCAP: Hierarchical Reinforcement Learning With Kinematic Constraints for Adaptive Parking3
The hierarchical reinforcement learning with kinematic constraints for adaptive parking (HRL-KCAP) is presented, capable of significantly optimizing path efficiency and enhancing parking precision.
- BLISS: A Lightweight Bilevel Influence Scoring Method for Data Selection in Language Model Pretraining3
BLISS is introduced, a lightweight data selection method that operates entirely from scratch, without relying on any external pretrained oracle models, while explicitly accounting for the long-term impact of selected data.
- Association prediction of lncRNAs and diseases using multiview graph convolution neural network3
This work proposes MVIGCN, a graph convolutional network (GCN)-based method integrating multimodal data to predict lncRNA-disease associations, and demonstrates improved reliability over single-view methods, highlighting its potential to identify disease-related lncRNA biomarkers.
- 3
- An Advanced Framework for Time-lapse Microscopy Image Analysis3
A robust and user-friendly pipeline to preprocess the images and track the behaviors of thousands of cancer cells simultaneously for a better drug response prediction of cancer patients is proposed.
- 3
- Regional Industrial Water Demand Prediction Based on Improved Series Gray Neural Network3
The results show that the improved prediction series gray neural network model has higher precision and is a practical, strong prediction method finally predicted Nanjing 2015 ~ 2016 industrial water demand.
- 3
- 3
- fast triangle rasterization using irregular z-buffer on cuda3
This work provides a complete description to the irregular Z-Buffer based shadow mapping software implementation on CUDA, capable of generating aliasing-free shadows at a throughput of dozens of million triangles per second.
- TDS-CLIP: Temporal Difference Side Network for Efficient Video Action Recognition2
A novel memory-efficient Temporal Difference Side Network (TDS-CLIP) is proposed to balance knowledge transferring and temporal modeling, avoiding backpropagation in frozen parameter models and introduces a Temporal Difference Adapter (TD-Adapter), which can effectively capture local temporal differences in motion features to strengthen the model's global temporal modeling capabilities.
- 2
- 2
- SLPDR: A Benchmark for Ship License Plate Detection and Recognition2
This paper proposes an SLP detection model named YOLO-SSA and evaluates this model as well as typical detection methods on the SLPDR dataset and demonstrates that the proposed YOLO-SSA achieves better SLP detection performance by enhancing the features where ships and SLPs are located.
- UIMM-Tracker: IMM-Based with Uncertainty Detection for Video Satellite Infrared Small-Target Tracking2
This paper proposes an infrared small-target tracking method, the UIMM-Tracker, based on the tracking-by-detection (TbD) paradigm, outperforming classical methods and demonstrating its effectiveness.
- 2
- 2
- 2
- 2
- A RDA-Based Deep Reinforcement Learning Approach for Autonomous Motion Planning of UAV in Dynamic Unknown Environments2
A novel reward difference amplifying (RDA) scheme is proposed to reshape the conventional reward functions and is introduced into state-of-the-art DRLs to constructs novel DRL algorithms for the planner’s learning.
- 2
- 2
- Some new inequalities for continuous fusion frames and fusion pairs2
Using operator methods, some new inequalities for continuous fusion frames and fusion pairs are established and these results extend and improve ones obtained by Balan, Casazza and Găvruţa.
- 2
- 2
- On the Dynamic Scheduling of Task Farm Patterns on a Heterogeneous CPU-GPGPU Environment2
A multi-round scheduling algorithm and a scheduling framework for farm-pattern based applications on a heterogeneous architecture and the experimental results demonstrate that the dynamic scheduling algorithm achieves better to similar performances as compared to some of the well-known scheduling algorithms for CPU-GPGPU systems.
- 1
- 1
- Heterophily-Aware Spectral GCN for Population-Level Brain Disorder Prediction1
A population graph with a phenotypic encoder, which leverages rs-fMRI and phenotypic data to model complex relationships between subjects and enables GNN to learn population-level features, and a heterophily-aware spectral graph convolution network that incorporates local similarity-based learning to assess node homophily and addresses the heterophily issue.
- Neuroimaging pattern interactions for suicide risk in depression captured by ensemble learning over transcriptome-defined parcellation1
An ensemble learning scheme over transcriptome-defined parcellations (TDP) to explore homogeneously parcellated brain patterns and their interactions displayed higher suicide predictive performance, relative to that over the regions level, and over null model.
- 1
- 1
- 1
- 1
- Development of a Locking and Releasing Mechanism for the Two-Dimensional Turntable of a Space Camera1
- 1
- 1
- The Automatic Operation and Maintenance Scheme for Cloud Data Centers in Multi-site Fusion1
A new scheme for automatic operation and maintenance is introduced, so as to meet the technical requirements of automaticoperation and maintenance, that can automatically manage resources and improve the efficiency for operation andtenance in cloud data centers.
- 1
- Exploring Brain Hemodynamic Response Patterns via Deep Recurrent Autoencoder1
The proposed DRAE model is to combine the deep recurrent neural network and autoencoder to automatically characterize the meaningful functional brain networks and corresponding diverse and complex hemodynamic response patterns underlying tfMRI data simultaneously.
- 1
- 1
- 1
- 1
- 1
- An Error Bound for the Sensor Scheduling Problem1
It is shown that under a given N -horizon sensor schedule σ ∈ M, the measurement sequence is determined by: y(t) = yσ(t)(t) + vσ(T)x(t), ∀t = 0, . . . , N.
- Discussion on digital architecture,part II:Evidence-based design1
The idea of evidence-based design is right the extension of BIM in the sense of all lifecycle information.
- Discussion on digital architecture,part I:termination1
The study of the dynamic information flow in BLM will make architecture design significant in the sense of EBD, which will give rise to a modern development in the industry system and ideas.
- Fault prediction of inertial component based on combined forecasting of ARIMA and RBF neural network1
A new method of combined predicting based on ARIMA forecast and RBF network is put forward, which shows that the combined model has better accuracy and it is an effective method for the fault prediction of inertial component.
- 1
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Topography High Sensitivity of Discriminative Fibers in ASD Children with Parcellation-Free Machine Learning Method–
A reliable classification model for early diagnosis for ASD is established that can output white matter fibers that contribute to classification at the fiber level, which may be meaningful for exploring the pathogenesis and potential intervention targets of ASD.
- –
- –
- –
- –
- –
- –
- Research on transformer load forecasting based on deep learning–
A short-term load prediction model based on long and short-term memory recurrent neural network (LSTM) with deep learning is built to predict the load level of transformers in the coming week using the historical load of transformers, holiday information and meteorological data as input.
- –
- Terahertz deep learning computed tomography in extreme sparse view–
A novel deep learning-based framework to decrease data acquisition time and enhance the 3D reconstruction resolution of sparse view THz-CT using asymmetric convolution blocks and channel attention mechanisms, which effectively suppresses terahertz image artifact in extreme sparse views.
- –
- –
- –
- –
- –
- –
- Super-resolution Reconstruction Based on Capsule Generative Adversarial Network–
Using the capsule network's excellent perception of hierarchical spatial information and local feature relationships, the author proposes a super-resolution reconstruction based on capsule network CSRGAN, which shows that compared with the pure convolution method RDN, the PSNR value ofCSRGAN is increased by 0.14, closer to the original image.
- Improving Non-Autoregressive Machine Translation via Autoregressive Training–
This paper introduces a multi-task framework to improve the performance of NAT task, and uses a fully shared encoder-decoder network to train NAT task and AT task simultaneously.
- –
- –
- Research on Assessment of E-learning Efficiency Based on Facial Direction–
An image processing technique based method to obtain index of distraction by estimating head pose of learner is proposed and shows that the success rate of judging distraction status can reach nearly 80% if the head pose estimation succeed.
- –
- Study on Shape Recognition Based on Contour–
A contour-based method of feature extraction and shape recognition is presented, and by using the features, a two-stake recursive algorithm for recognition is proposed.
- –
- Learning Affine Robust Binary Codes Based on Locality Preserving Hash–
LPH preserves local structure by embedding data into a sub-manifold, and performing binarization that minimize false classification ratio while keeping partition balanced, and experiments show that LPH is easy to train and performs better than state-of-the-art methods with more compact binary codes.
- –
- –
- Description of Image Boundary and Spatial Relation Based on Eigenvalue Decomposition–
Analysis and experiments prove that the spatial relation descriptor is robust under translation, scaling,rotation and small noise, and it has good similarity discriminating ability.
- Graph user interface automatic testing model based on software monitoring–
In order to improve the coverage and efficiency of locating faults and defect the faults in running repetitiously in Graph User Interface (GUI) automatic testing, a GUI automated testing model based on software monitoring was proposed.
- Task Planning Based on the Description Logic of Tasks in Joint Operation Scenarios–
A method named TPDLT is proposed, which provides a formal description of the task planning in joint operation simulations based on the description logic of tasks, and gives a definition of the completeness of task plans and a decidable theorem of the completion under joint strategies.
- VideoMap: Interactive Search System by MCG-ICT-CAS*–
This paper describes the highlights of the interactive search system VideoMap for TRECVID 2009, which has a map based displaying interface, which gives the user a global view about the similarity relationships among the whole video collection, and provides an active annotating manner to quickly localize the potential positive samples.
- –
- Design for Real-time Multi-task Scheduling Method–
A real-time multi-task scheduling method, taking the key and value-density of task as main part and adding the constraint relations among tasks, named key—value-density—task constraint is proposed, and the design rule of priority is given.
- –
- A new method for reliability estimation by using prior information–
A new Bayes analysis routine is put forward, which can get rid of the difficulties in prior distribution determination for multi-parameters model, and is mainly used for reliability estimation ofMulti-parameter life distribution.
- Application of Fuzzy Neural Network to Multi-Faults Diagnosis of Missile Power System–
The result of fault diagnosis instance shows that fuzzy neural network based on differential evolution algorithm is fast and can classify accurately single and multi fault pattern of missile power system.
- –
- The Study on the Self-adjustment Method for Improving the Brain Fatigue of the Students–
College students at schools could achieve the target that they keep the body healthy and spirit full of vigor, and they are in a cheerful frame by improving the intelligence of the college students so as to enable the collegeStudents to finish their studies.
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 2026-10-12. 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.