Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks345 from public data
- Effective Whole-body Pose Estimation with Two-stages Distillation446
This work presents a two-stage pose Distillation for Whole-body Pose estimators, named DWPose, to improve their effectiveness and efficiency and releases a series of models with different sizes, from tiny to large, for satisfying various downstream tasks.
- Focal and Global Knowledge Distillation for Detectors432
Focal and Global Distillation (FGD) is proposed, which separates the foreground and background, forcing the student to focus on the teacher's critical pixels and channels, and compensating for missing global information in focal distillation.
- Self-Attention ConvLSTM for Spatiotemporal Prediction384
A novel self-attention memory (SAM) is proposed to memorize features with long-range dependencies in terms of spatial and temporal domains and is applied to perform frame prediction on the MovingMNIST and KTH datasets and traffic flow Prediction on the TexiBJ dataset.
- 265
- Differentially Private Data Release through Multidimensional Partitioning193
This paper proposes two multidimensional partitioning strategies including a baseline cell-based partitioning and an innovative kd-tree based partitioning for differentially private histogram release based on an interactive differential privacy interface.
- Automated known problem diagnosis with event traces188
This work proposes to use system behavior information such as system event traces to build correlations with solved problems, instead of using only vague text descriptions as in existing practices to enable automatic identification of the root cause of a problem if it is a known one, which would further lead to its resolution.
- Strider: a black-box, state-based approach to change and configuration management and support184
Troubleshooting of configuration failures are used to demonstrate that the Strider approach reduces problem complexity by several orders of magnitude, making root-cause analysis possible.
- From Knowledge Distillation to Self-Knowledge Distillation: A Unified Approach with Normalized Loss and Customized Soft Labels166
This work unifies the formulations of the two tasks by decomposing and reorganizing the generic KD loss into a Normalized KD (NKD) loss and customized soft labels for both target class (image’s category) and non-target classes named Universal Self-KD (USKD).
- ColorNet: Investigating the Importance of Color Spaces for Image Classification160
This model essentially takes an RGB image as input, simultaneously converts the image into 7 different color spaces and uses these as inputs to individual densenets to reduce computation overhead and number of hyperparameters required.
- Low-Rank Preserving Projections158
This paper proposes a novel dimensionality reduction method, named low-rank preserving projections (LRPP) for image classification, and shows the effectiveness and the feasibility of the proposed method with encouraging results.
- 147
- FlatQuant: Flatness Matters for LLM Quantization144
This paper proposes FlatQuant (Fast and Learnable Affine Transformation), a new post-training quantization approach that enhances the flatness of weights and activations, and identifies optimal affine transformations for each linear layer, calibrated in hours via a lightweight objective.
- ChartBench: A Benchmark for Complex Visual Reasoning in Charts127
ChartBench, a comprehensive benchmark designed to assess chart comprehension and data reliability through complex visual reasoning, is proposed and two baselines based on the chain of thought and supervised fine-tuning are proposed to improve model performance on unannotated charts.
- Convolution Meets LoRA: Parameter Efficient Finetuning for Segment Anything Model121
Conv-LoRA not only preserves SAM's extensive segmentation knowledge but also revives its capacity of learning high-level image semantics, which is constrained by SAM's foreground-background segmentation pretraining.
- CricaVPR: Cross-Image Correlation-Aware Representation Learning for Visual Place Recognition115
A robust global representation method with cross-image correlation awareness for VPR, named CricaVPR, which uses the attention mechanism to correlate multiple images within a batch and introduces the multi-scale local information to further enhance the cross-image correlation-aware representation.
- DreamDiffusion: Generating High-Quality Images from Brain EEG Signals111
Quantitative and qualitative results demonstrate the effectiveness of the proposed DreamDiffusion method as a significant step towards portable and low-cost ``thoughts-to-image'', with potential applications in neuroscience and computer vision.
- Towards Seamless Adaptation of Pre-trained Models for Visual Place Recognition100
A hybrid adaptation method to achieve both global and local adaptation efficiently, in which only lightweight adapters are tuned without adjusting the pre-trained model, and a mutual nearest neighbor local feature loss, which ensures proper dense local features are produced for local matching and avoids time-consuming spatial verification in re-ranking.
- Learning to Run Challenge Solutions: Adapting Reinforcement Learning Methods for Neuromusculoskeletal Environments98
This work presents eight solutions that used deep reinforcement learning approaches, based on algorithms such as Deep Deterministic Policy Gradient, Proximal Policy Optimization, and Trust Region Policyoptimization, to make it run as fast as possible through an obstacle course.
- 90
- SparseFlex: High-Resolution and Arbitrary-Topology 3D Shape Modeling89
This paper introduces SparseFlex, a novel sparse-structured isosurface representation that enables differentiable mesh reconstruction at resolutions up to 10243 directly from rendering losses, and introduces a frustum-aware sectional voxel training strategy that activates only relevant voxels during rendering, dramatically reducing memory consumption and enabling high-resolution training.
- Structurally Incoherent Low-Rank Nonnegative Matrix Factorization for Image Classification86
This paper proposes a novel nonnegative factorization method, called structurally incoherent low-rank NMF (SILR-NMF), in which they jointly consider structural incoherence and low- rank properties of data for image classification.
- Quantization Hurts Reasoning? An Empirical Study on Quantized Reasoning Models83
This paper conducts the first systematic study on quantized reasoning models, evaluating the open-sourced DeepSeek-R1-Distilled Qwen and LLaMA families ranging from 1.5B to 70B parameters, QwQ-32B, and Qwen3-8B using state-of-the-art algorithms at varying bit-widths.
- Learning Imbalanced Data with Vision Transformers80
This paper systematically investigates the ViTs' performance in LTR and proposes LiVT to train ViTs from scratch only with LT data, and derives the unbiased extension of Sigmoid and compensates extra logit margins for deploying it to ameliorate its performance.
- Uni-Perceiver v2: A Generalist Model for Large-Scale Vision and Vision-Language Tasks75
This paper proposes Uni-Perceiver v2, which is the first generalist model capable of handling major large-scale vision and vision-language tasks with competitive performance and proposes an effective optimization technique named Task-Balanced Gradient Normalization to ensure stable multi-task learning with an unmixed sampling strategy, helpful for tasks requiring large batch-size training.
- UPop: Unified and Progressive Pruning for Compressing Vision-Language Transformers71
The UPop framework is proposed as a universal vison-language Transformer compression framework, which incorporates unifiedly searching multimodal subnets in a continuous optimization space from the original model, which enables automatic assignment of pruning ratios among compressible modalities and structures.
- 70
- PS-NeRV: Patch-Wise Stylized Neural Representations for Videos69
This work proposes a patch-wise solution to represent a video with implicit neural representations, PS-NeRV, which represents videos as a function of patches and the corresponding patch coordinate, and achieves excellent reconstruction performance with fast decoding speed.
- 68
- 66
- Low-Rank 2-D Neighborhood Preserving Projection for Enhanced Robust Image Representation64
A novel NPP method called low-rank 2DNPP (LR-2DNPP) is proposed, which encodes the structural incoherence of the learned clean data to enhance the discriminative ability for feature extraction and uses seven public image databases to verify the performance of the proposed methods.
- SWEM: Towards Real-Time Video Object Segmentation with Sequential Weighted Expectation-Maximization61
A novel Sequential Weighted Expectation-Maximization (SWEM) network to greatly reduce the redundancy of memory features and maintains a fixed number of template features in memory, which ensures the stable inference complexity of the VOS system.
- Nonnegative Discriminant Matrix Factorization60
A novel method called nonnegative discriminant matrix factorization (NDMF) is proposed for image classification that integrates the nonnegative constraint, orthogonality, and discriminant information in the objective function and is proposed to enhance the discriminant ability of the learned base matrix.
- Make Encoder Great Again in 3D GAN Inversion through Geometry and Occlusion-Aware Encoding58
This work introduces a novel encoder-based inversion framework based on EG3D, one of the most widely-used 3D GAN models, and uses the inherent properties of EG3D’s latent space to design a discriminator and a background depth regularization.
- 55
- 53
- Distilling Semantic Priors from SAM to Efficient Image Restoration Models52
This work proposes a general framework to distill SAM's semantic knowledge to boost exiting IR models without interfering with their inference process, and designs a semantic-guided relation (SGR) module for SPD, which ensures semantic feature representation space consistency to fully distill the priors.
- HyP 2 Loss: Beyond Hypersphere Metric Space for Multi-label Image Retrieval51
A novel metric learning framework with Hybrid Proxy-Pair Loss (HyP$^2$ Loss) that constructs an expressive metric space with efficient training complexity and focuses on optimizing the hypersphere space by learnable proxies and excavating data-to-data correlations of irrelevant pairs.
- ViTKD: Practical Guidelines for ViT feature knowledge distillation50
This paper designs a series of controlled experiments and derives three practical guidelines for ViT's feature distillation and proposes the feature-based method ViTKD which brings consistent and considerable improvement to the student.
- 49
- 49
- Handwritten Character Recognition Using Orientation Quantization Based on 3D Accelerometer47
This system could be used to reduce the size of handheld devices by discarding number keys and make human computer interaction more convenient and interesting.
- 46
- High-fidelity Facial Avatar Reconstruction from Monocular Video with Generative Priors45
This work proposes an efficient method to construct the personalized generative prior based on a small set of facial images of a given individual that allows for photo-realistic rendering with novel views, and the face reenactment can be realized by performing navigation in the latent space of 3D-GAN.
- When Noisy Labels Meet Long Tail Dilemmas: A Representation Calibration Method45
This paper proposes a representation calibration method RCAL, which assumes that without incorrect labeling and class imbalance, the representations of instances in each class conform to a multivariate Gaussian distribution, which is much milder and easier to be checked.
- One Model to Edit Them All: Free-Form Text-Driven Image Manipulation with Semantic Modulations44
The proposed Free-Form CLIP has a cross-modality semantic modulation module containing semantic alignment and injection, which performs the automatic latent mapping via linear transformations with a cross attention mechanism and can leverage text prompts with multiple semantic meanings for image manipulation.
- 42
- 42
- 39
- 37
- Friends Troubleshooting Network: Towards Privacy-Preserving, Automatic Troubleshooting37
This paper constructs Friends Troubleshooting Network (FTN), a peer-to-peer overlay network, where the links between peer machines reflect the friendship of their owners, and uses historyless and futureless random-walk in the FTN to preserve privacy.
- A fast minimal path active contour model37
The proposed minimal path active contour model speeds up the search and reduces the "metrication error" frequently encountered in the classical graph search methods e.g., the dynamic programming minimal path (DPMP) method.
- HiFace: High-Fidelity 3D Face Reconstruction by Learning Static and Dynamic Details36
The static detail is modeled as the linear combination of a displacement basis and the dynamic detail is modeled as the linear interpolation of two displacement maps with polarized expressions, which enable HiFace to reconstruct high-fidelity 3D shapes with animatable details.
- Fast Registration for Cross-Source Point Clouds by using Weak Regional Affinity and Pixel-Wise Refinement36
Two components, the weak region affinity and pixel-wise refinement, are proposed to maintain the global and local information of 3D point clouds and integrated into an iterative tensor-based registration algorithm to solve the cross-source point cloud registration problem.
- Nuclear Norm-Based 2DLPP for Image Classification35
A novel locality preserving projection method called nuclear norm-based two-dimensional locality preserving projections (NN-2DLPP), which recovers the noisy data matrix through low-rank learning and removes noise in data, and keeps data points belonging to the same class as close to each other as possible in the new projective subspace.
- CustomNet: Zero-shot Object Customization with Variable-Viewpoints in Text-to-Image Diffusion Models34
The CustomNet is introduced, a novel object customization approach that explicitly incorporates 3D novel view synthesis capabilities into the object customization process, and facilitates zero-shot object customization without test-time optimization, offering simultaneous control over the viewpoints, location, and background.
- Why is the Winner the Best?34
A multicenter study with all 80 benchmarking competitions that were conducted in the scope of IEEE ISBI 2021 and MICCAI 2021 revealed common characteristics of winning solutions: the reflection of the metrics in the method design and the focus on analyzing and handling failure cases.
- A Coarse-to-Fine Algorithm for Registration in 3D Street-View Cross-Source Point Clouds34
This paper proposes a effective coarse-to-fine algorithm to detect and register a small scale SFM point cloud in a large scale Lidar point cloud and can make a contribution to many applications, such as robotics and smart city development.
- Resampling on a Pseudohexagonal Grid34
This paper first describes the options and procedures for constructing such a pseudo-hexagonal grid and then demonstrates techniques of resampling digital images on the pseudo hexagonal grid.
- 32
- TextIR: A Simple Framework for Text-Based Editable Image Restoration30
A sophisticated framework enabling users to guide the restoration of deteriorated images via textual descriptions via textual descriptions is developed, and the text-image compatibility feature of CLIP enhances the integration of textual and visual data.
- 30
- Projective robust nonnegative factorization30
A framework named Projective Robust Nonnegative Factorization (PRNF) is proposed in this paper for robust image feature extraction and classification and results are compared with state-of-the-art robust NMF-based algorithms.
- 30
- Feature Augmented Memory with Global Attention Network for VideoQA29
A novel videoQA framework which progressively refines the representations of videos and questions from fine to coarse grain in a sequence-sensitive manner and achieves state-of-the-art performance on VideoQA benchmark datasets.
- 28
- 28
- 28
- Perceive, Understand and Restore: Real-World Image Super-Resolution with Autoregressive Multimodal Generative Models27
Experimental results demonstrate that PURE preserves image content while generating realistic details, especially in complex scenes with multiple objects, showcasing the potential of autoregressive multimodal generative models for robust Real-ISR.
- Reducing the Annotation Effort for Video Object Segmentation Datasets27
A deep convolutional network is used to automatically create pseudo-labels on a pixel level from much cheaper bounding box annotations and it is found that adding a manually annotated mask in only a single video frame for each object is sufficient to generate pseudo-Labels which can be used to train a VOS method to reach almost the same performance level as when training with fully segmented videos.
- Deep Homography Estimation for Visual Place Recognition25
A transformer-based deep homography estimation (DHE) network that takes the dense feature map extracted by a backbone network as input and fits homography for fast and learnable geometric verification is proposed and can outperform several state-of-the-art methods.
- SuperVLAD: Compact and Robust Image Descriptors for Visual Place Recognition24
Experimental results suggest that, when paired with a transformer-based backbone, the SuperVLAD shows better domain generalization performance than NetVLAD with significantly fewer parameters, and surpasses state-of-the-art methods with lower feature dimensions on several benchmark datasets.
- Semantic-Sparse Colorization Network for Deep Exemplar-Based Colorization24
Semantic-Sparse Colorization Network (SSCN) is proposed to transfer both the global image style and detailed semantic-related colors to the gray-scale image in a coarse-to-fine manner and can perfectly balance the global and local colors while alleviating the ambiguous matching problem.
- Self-Adaptive Double Bootstrapped DDPG24
SOUP improves the efficiency of exploration by multiple actor heads capturing more potential actions and multiple critic heads evaluating more reasonable Q-values collaboratively and achieves faster learning by at least 45% while improving cumulative reward and stability substantially in comparison to vanilla DDPG on OpenAI Gym's MuJoCo environments.
- 23
- 22
- Rethinking Knowledge Distillation via Cross-Entropy21
This work decomposes the KD loss to explore its relation with the original cross-entropy (CE) loss, and proposes a new KD loss (NKD), which achieves state-of-the-art performance on CIFAR-100 and ImageNet.
- 21
- Weakly Supervised Temporal Action Localization via Dual-Prior Collaborative Learning Guided by Multimodal Large Language Models20
A novel learning paradigm termed MLLM4WTAL is introduced, which harnesses the potential of MLLM to offer temporal action key semantics and complete semantic priors for conventional Weakly-supervised Temporal Action Localization (WTAL) methods.
- SelaVPR++: Towards Seamless Adaptation of Foundation Models for Efficient Place Recognition20
An extension of the SelaVPR, called SelaVPR++, is proposed that is highly efficient in training time, GPU memory usage, and retrieval latency, as well as outperforms the state-of-the-art methods by a large margin.
- Learning to Learn from APIs: Black-Box Data-Free Meta-Learning20
A Bi-level Data-free Meta Knowledge Distillation (BiDf-MKD) framework to transfer more general meta knowledge from a collection of black-box APIs to one single meta model by just querying APIs is proposed.
- HOSE-Net: Higher Order Structure Embedded Network for Scene Graph Generation20
A novel structure-aware embedding-to-classifier(SEC) module to incorporate both local and global structural information of relationships into the output space and a hierarchical semantic aggregation module to reduces the number of subspaces by introducing higher order structural information.
- TreeNet: Learning Sentence Representations with Unconstrained Tree Structure20
A novel neural network, namely TreeNet, to capture sentences structurally over the raw unconstrained constituency trees, where the number of child nodes can be arbitrary, which proves to be a simple and efficient mechanism for semantic synthesis.
- 20
- Efficient Conditional Diffusion Model with Probability Flow Sampling for Image Super-resolution19
This paper designs a continuous-time conditional diffusion model for image super-resolution, which enables the use of probability flow sampling for efficient generation and designs an image quality loss as a complement to the score matching loss of diffusion models, further improving the consistency and quality of super-resolution.
- CMS-LSTM: Context Embedding and Multi-Scale Spatiotemporal Expression LSTM for Predictive Learning19
This work designs CMS-LSTM to focus on context correlations and multi-scale spatiotemporal flow with details on fine-grained locals, containing two elaborate de-signed blocks: Context Embedding and Spatiotem temporal Expression blocks.
- 19
- Supervised Fine-tuning in turn Improves Visual Foundation Models18
A two-stage method ViSFT (Vision SFT) is proposed to unleash the fine-grained knowledge of vision foundation models and shows improvements across various out-of-domain benchmarks including vision and vision-linguistic scenarios.
- Facial Expression Recognition with Multi-scale Convolution Neural Network18
A triplet-loss learning method is used to train the architecture, which improves the discrimination of deep features and achieves superior performance to other state-of-the-art methods on the FER2013 dataset.
- A Quality-Controllable Encryption for H.264/AVC Video Coding18
The goal has been to provide an efficient way to scramble the video streams to prevent illegal users from plagiarizing and to ensure that even the deepest scrambling level adds trivial performance overhead to the standard H.264 encoding/decoding process.
- ComRoPE: Scalable and Robust Rotary Position Embedding Parameterized by Trainable Commuting Angle Matrices17
This work proposes ComRoPE, which generalizes RoPE by defining it in terms of trainable commuting angle matrices, and demonstrates that pairwise commutativity of these matrices is essential for RoPE to achieve scalability and positional robustness.
- AANet: Aggregation and Alignment Network with Semi-hard Positive Sample Mining for Hierarchical Place Recognition17
A Dynamically Aligning Local Features (DALF) algorithm to align the local features under spatial constraints is proposed and presented, which is significantly more efficient than the methods that need geometric consistency verification.
- Towards Effective Collaborative Learning in Long-Tailed Recognition17
This article observes that the knowledge transfer between experts is imbalanced in terms of class distribution, which results in limited performance improvement of the minority classes, and proposes a re-weighted distillation loss by comparing two classifiers' predictions, which are supervised by online distillation and label annotations.
- Graph Cuts Stereo Matching Based on Patch-Match and Ground Control Points Constraint17
A new method that integrates Patch-Match and graph cuts GC is proposed in order to achieve good results in both complex and low texture regions.
- Proxy+: Simple Proxy Augmentation for Dynamic Content Processing17
This paper proposes an approach to enable dynamic content caching at enhanced Web proxies which requires only simple modifications to existing applications.
- EDTformer: An Efficient Decoder Transformer for Visual Place Recognition16
This work proposes an Efficient Decoder Transformer (EDTformer) for feature aggregation, which consists of several stacked simplified decoder blocks followed by two linear layers to directly produce robust and discriminative global representations.
- 16
- 16
- The Dorsal Anterior Cingulate Cortex Modulates Dialectical Self-Thinking16
The data showed that activity in the dACC during the processing of self-relevant conflicting information was positively correlated with participants’ dispositional level of naïve dialecticism (measured with the Dialectical Self Scale), which supports the hypothesis that the d ACC has a key role in dialectical self-thinking.
- 16
- IMDPrompter: Adapting SAM to Image Manipulation Detection by Cross-View Automated Prompt Learning15
This work develops a cross-view prompt learning paradigm called IMDPrompter, which no longer relies on manual guidance, enabling automated detection and localization of image manipulation detection and proposes components such as Cross-view Feature Perception, Optimal Prompt Selection, and Cross-View Prompt Consistency, which facilitate cross-view perceptual learning.
- 15
- Semantic Distillation from Neighborhood for Composed Image Retrieval15
This work proposes a novel framework termed SemAntic Distillation from Neighborhood (SADN) for composed image retrieval that significantly outperforms baseline methods in terms of results and ablation studies on CIRR and FashionIQ.
- Task Groupings Regularization: Data-Free Meta-Learning with Heterogeneous Pre-trained Models15
It is found that model heterogeneity introduces a heterogeneity-homogeneity trade-off, where homogeneous models reduce task conflicts but also increase the overfitting risk, which is crucial for learning shared representations across tasks.
- 15
- Implementing DRM over Peer-to-Peer Networks with Broadcast Encryption15
This paper proposes a strategy to incorporate DRM mechanism into P2P network architecture and construct a system which ensures efficient content sharing as well as reliable media protection and demonstrates that the new strategy outperforms traditional solutions on alleviating communication overhead of License Server, minishing peer latency of authorization, and importing security modules without much modification to original peer-to-peer infrastructure.
- Composite Chaotic Pseudo-Random Sequence Encryption Algorithm for Compressed Video15
A composite chaotic cryptography scheme was developed to encrypt the compressed video with the logistics map with a Z(2 31 -1) field linear congruential algorithm to strengthen the security of the mono-chaotic cryptography.
- Free: Faster and Better Data-Free Meta-Learning14
The Faster and Better Data-Free Meta-Learning (FREE) framework is introduced, which contains a meta-generator for rapidly recovering training tasks from pre-trained models and a meta-learner for generalizing to new unseen tasks.
- 14
- Weighted-PCANet for Face Recognition14
Weighted-PCANet, a novel feature learning method is proposed to face recognition by combining Linear Regression Classification model (LRC) and PCANet construction, which outperform than various mainstream methods including PCANnet for face recognition on Extended YaleB dataset.
- Architecture, Dataset and Model-Scale Agnostic Data-free Meta-Learning13
A unified framework, dubbed PURER, is proposed, which contains a unified framework for meta training with ECI as an adversarial form in an end-to-end manner and a simple plug-and-play supplement—ICFIL—only used during meta testing to narrow the gap between meta training and meta testing task distribution.
- Privacy-Preserving Model Upgrades with Bidirectional Compatible Training in Image Retrieval13
This work proposes a new model upgrade paradigm, termed Bidirectional Compatible Training (BiCT), which will upgrade the old gallery embeddings by forward-compatible training towards the embedding space of the backward-compatible new model.
- 13
- Diverse Image Inpainting with Normalizing Flow13
Flow-Fill is proposed, a novel two-stage image inpainting framework that utilizes a conditional normalizing flow model to generate diverse structural priors in the first stage and achieves real-time inference speed and eliminates pixel by pixel errors.
- TranSlider: Transfer Ensemble Learning from Exploitation to Exploration13
The concept of transfer ensemble learning is introduced, a new direction to tackle the over-fitting of transfer strategies, and the proposed Transferability Slider (TranSlider) achieves the state-of-the-art on comprehensive benchmark datasets.
- Fast Mode and Depth Decision Algorithm for Intra Prediction of Quality SHVC13
A fast mode and depth decision algorithm for Intra prediction of Quality SHVC that could improve coding speed by 61.31% on average with negligible coding efficiency losses.
- 12
- VQRAE: Representation Quantization Autoencoders for Multimodal Understanding, Generation and Reconstruction12
VQRAE, a Vector Quantization version of Representation AutoEncoders, is proposed, which pioneers the first exploration in unified representation to produce Continuous semantic features for image understanding and Discrete tokens for visual generation within a unified tokenizer.
- 12
- ChartMoE: Mixture of Diversely Aligned Expert Connector for Chart Understanding12
This work proposes ChartMoE, which employs the mixture of expert (MoE) architecture to replace the traditional linear pro-jector to bridge the modality gap and trains multiple linear connectors through distinct alignment tasks, which are utilized as the foundational initialization parameters for different experts.
- 12
- Prune&Comp: Free Lunch for Layer-Pruned LLMs via Iterative Pruning with Magnitude Compensation11
It is identified that removing any layer induces a magnitude gap in hidden states, and it is demonstrated that a simple compensation operation leads to superior performance in iterative layer pruning.
- DFVSR: Directional Frequency Video Super-Resolution via Asymmetric and Enhancement Alignment Network11
This paper reconsider object motion from a new perspective and proposes Directional Frequency Representation (DFR), which not only borrows the property of frequency representation of detail and structure information but also contains the direction information of the object motion that is extremely significant in videos.
- 11
- Self-Supervised Mixture-of-Experts by Uncertainty Estimation11
This paper proposes SelfSupervised Mixture-of-Experts (SUM), an effective algorithm driven by predictive uncertainty estimation for multitask RL that learns faster and achieves better performance by efficient transfer and robust generalization, outperforming several related methods on extended OpenAI Gym’s MuJoCo multi-task environments.
- Automatic Image Annotation Using Adaptive Weighted Distance in Improved K Nearest Neighbors Framework11
An adaptive weighted distance method which incorporates the CNN convolutional neural network feature and multiple handcrafted features is proposed to handle the label-image-matching and label-imbalance issues, while the K nearest neighbors framework is improved by using the neighborhood with all labels which can reduce the effects of thelabel-missing problem.
- 11
- 11
- 11
- Chaos-based encryption algorithm for compressed video11
A chaotic selective encryption of compressed video (CSECV) which exploits the characteristics of the compressed video and has significant advantages over existing algorithms for security, decryption speed, implementation flexibility, and error preservation.
- 11
- 11
- Cobra: Efficient Line Art COlorization with BRoAder References10
Cobra is an efficient and versatile method that supports color hints and utilizes over 200 reference images while maintaining low latency, and leverages specially designed positional encodings, causal sparse attention, and Key-Value Cache to effectively manage long-context references and ensure color identity consistency.
- CustomNet: Object Customization with Variable-Viewpoints in Text-to-Image Diffusion Models10
CustomNet is presented, a unified encoder-based object customization framework that explicitly incorporates 3D novel view synthesis capabilities into the customization process, and outperforms other customization methods regarding identity preservation, diversity, and harmony.
- Enhanced Image Deblurring: An Efficient Frequency Exploitation and Preservation Network10
This paper proposes a novel Efficient Frequency Exploitation and Preservation Network (EFEP), which outperforms state-of-the-art models on both quantitative and qualitative evaluations.
- 9
- 9
- 9
- 9
- 9
- ChartPoint: Guiding MLLMs with Grounding Reflection for Chart Reasoning8
This paper proposes PointCoT, which integrates reflective interaction into chain-of-thought reasoning in charts by prompting MLLMs to generate bounding boxes and re-render charts based on location annotations, and establishes connections between textual reasoning steps and visual grounding regions.
- 8
- 8
- Rethink Long-Tailed Recognition with Vision Transforms8
This paper revisits recent LTR methods with promising Vision Transformers and proposes the Predictive Distribution Calibration (PDC) as a novel metric for LTR, where the model tends to simply classify inputs into common classes.
- ITstyler: Image-optimized Text-based Style Transfer8
This work converts text input to the style space of the pre-trained VGG network to realize a more effective style swap and uses CLIP's multi-modal embedding space to learn the text-to-style mapping with the image dataset only.
- Cross-modal Representation Learning and Relation Reasoning for Bidirectional Adaptive Manipulation8
This paper introduces a robust solution for CAM, which includes two essential modules, namely Heterogeneous Representation Learning (HRL) and Cross-modal Relation Reasoning (CRR), designed to perform representation learning for cross- modal semantic alignment on heterogeneous graph nodes.
- 8
- Depth map super-resolution via low-resolution depth guided joint trilateral up-sampling8
A joint trilateral upsampling model is proposed to fuse the projected spatial distance measured from the LR depth map, the intensity variance extracted from the associated color image and theHR depth map generated in the last iteration to refine the HR depth map iteratively.
- 8
- 8
- Solving Continual Offline Reinforcement Learning with Decision Transformer7
This work compares AC-based offline algorithms with DT in the CORL framework and introduces multi-head DT (MH-DT) and low-rank adaptation DT (LoRA-DT) to mitigate DT's forgetting problem and demonstrates enhanced learning capabilities and superior memory efficiency.
- ALoRE: Efficient Visual Adaptation via Aggregating Low Rank Experts7
ALoRE is proposed, a novel PETL method that reuses the hypercomplex parameterized space constructed by Kronecker product to Aggregate Low Rank Experts using a multi-branch paradigm, disentangling the learned cognitive patterns during training.
- 7
- Neural Machine Translation with Dynamic Graph Convolutional Decoder7
This paper proposes a customized Dynamic Spatial-Temporal Graph Convolutional Decoder (Dyn-STGCD), which is designed for consuming source feature representations and their syntactic graph, and auto-regressively generating the target syntacticgraph and tokens simultaneously.
- HMC: Hierarchical Mesh Coarsening for Skeleton-free Motion Retargeting7
The qualitative results show that the proposed mesh-coarsening module is significantly helpful in preserving the moving consistency of different body parts on the target character due to disentangling body-part structures and mesh details in a hierarchical way.
- 7
- Efficient Multi-level Correlating for Visual Tracking7
This paper proposes a multi-level CF-based tracking approach named MLCFT which further explores the potential capacity of CF with two-stage detection: primal detection and oriented re-detection and demonstrates that this approach outperforms the most state-of-the-art trackers while tracking at speed of exceeded 16 frames per second on challenging benchmarks.
- 7
- 7
- Aligning Composed Query with Image via Discriminative Perception from Negative Correspondences6
To encourage awareness of the differences between matched and mismatched correspondences, DIPNEC introduces optimal transport with semantic preservation for reassignments on hard negative triplets and has demonstrated competitive performances on the experimental results and ablation studies on widely-used datasets FashionIQ and CIRR.
- Task-Distributionally Robust Data-Free Meta-Learning6
This paper proposes a trustworthy DFML framework comprising three components: synthetic task reconstruction, meta-learning with task memory interpolation, and automatic model selection, which seamlessly incorporates an automatic model selection mechanism to automatically filter out untrustworthy models during the meta-learning process.
- Adaptive Contrastive Learning for Learning Robust Representations under Label Noise6
This work proposes a novel adaptive contrastive learning method working at the pair level to select contrastive pairs adaptively, and considers the model's learning status to adjust the confidence threshold in a self-adaptive manner instead of fixing it.
- 6
- Design of Fatigue Driving Behavior Detection Based on Circle Hough Transform6
The method mainly identifies fatigue based on the detected target's face, head gestures, mouth aspect ratio (MAR), and eye condition, and then triggers an alarm through an intelligent auxiliary device, thereby fulfilling the purpose of alerting a group of high-risk drivers and preventing them from engaging in high- risk activities in an involuntary state.
- 6
- 6
- Tackling Instance-Dependent Label Noise with Dynamic Distribution Calibration6
This paper hypothesizes that, before training data are corrupted by label noise, each class conforms to a multivariate Gaussian distribution at the feature level, and proposes two methods based on the mean and covariance of multivariateGaussian distribution respectively.
- Structural Supervision for Word Alignment and Machine Translation6
This work proposes to incorporate the syntactic structure of both source and target tokens into the encoder-decoder framework, tightly correlating the internal logic of word alignment and machine translation for multi-task learning.
- Deep Interactive Video Inpainting: An Invisibility Cloak for Harry Potter6
This is the first deep learning-based interactive video inpainting framework that only uses a free form of user input as guidance (i.e. scribbles) instead of mask annotations, which has academic, entertainment, and commercial value.
- HAF-SVG: Hierarchical Stochastic Video Generation with Aligned Features6
This paper proposes a hierarchical recurrent VAE with a feature aligner, which can not only relax the independence assumption in typical VAE but also use a feature aligner to enable the decoder to obtain the aligned spatial information from the last observed frames.
- 6
- 6
- A fast mode decision algorithm applied to Coarse-Grain quality Scalable Video Coding6
A fast mode decision algorithm is proposed for the enhancement layer of MGS based on its specific coding structure and can achieve an average time saving of up to 85.44% with strong robustness and negligible loss in coding efficiency.
- 6
- Implementing Digital Right Management in P2P Content Sharing System6
This work proposes a new system based on P2P architecture for file sharing, to which it applies new DRM (Digital Right Management) mechanisms, and results have shown that the system outperforms other systems in features of security, server load and scalability.
- Identification of Chinese Unknown Word Based on Decision Tree6
This article suggests the recognition of unknown words as a question of classification first, that is, the segmentation fragments are divided into two categories as "combination" and "segregation" and decision tree is used to solve this problem of classification.
- 5
- Efficient Multi-Goal Reinforcement Learning via Value Consistency Prioritization5
This work proposes a novel value consistency prioritization (VCP) method, where the priority of samples is determined by the consistency of ensemble Q-values, and demonstrates that VCP achieves significantly higher sample efficiency than existing algorithms on a range of challenging goal-conditioned manipulation tasks.
- Modernn: Towards Fine-Grained Motion Details for Spatiotemporal Predictive Learning5
This paper carefully design Detail Context Block (DCB) to extract fine-grained details and improve the isolated correlation between upper context state and current input state and introduces Motion Details RNN (MoDeRNN), which outperforms existing state-of-the-art techniques qualitatively and quantitatively with lower computation loads.
- 5
- 5
- 5
- A Self-Supervised Feature Map Augmentation (FMA) Loss and Combined Augmentations Finetuning to Efficiently Improve the Robustness of CNNs5
This work proposes a new regularization loss called feature-map augmentation (FMA) loss which can be used during finetuning to make a model robust to several distortions in the input, and uses it to improve an existing state-of-the-art method called stability training (ST).
- 5
- 5
- 5
- 2D/3D Web Visualization on Mobile Devices5
This paper presents what was learned on engineering 3D Web visualization on both high-end and low-end mobile devices as the MWeb3D framework, which forms a distributed pipeline that move intensive computation from the mobile devices to server systems.
- 4
- 4
- 4
- Unlocking Tuning-Free Few-Shot Adaptability in Visual Foundation Models by Recycling Pre-Tuned LoRAs4
This work explores the potential of reusing diverse pre-tuned LoRAs without accessing their original training data, to achieve tuning-free few-shot adaptation in VFMs and distills a meta-LoRA from diverse pre-tuned LoRAs with a meta-learning objective, using surrogate data generated inversely from pre-tuned LoRAs themselves.
- 4
- 4
- 4
- FANet: Factor Analysis Neural Network4
A cascaded factor analysis network is proposed, which is suitable for extracting distributed semantic representations to various problems ranging from digit recognition and image classification to face recognition and combines a statistical methodfactor analysis with neural networks for excellent automatically learning ability and abundant semantic information.
- A Very Deep Sequences Learning Approach for Human Action Recognition4
This work proposes an approach based on CNNs and RNN-like models which have abilities to extract spatial and temporal features both, a CNN model can get static scores, a LSTM or GRU layer which gets dynamic class scores of human action.
- Fast Intra prediction algorithm for quality scalable video coding4
A new fast Intra coding algorithm for QSVC is proposed based on the coding structure of QSVC, temporal, spatial and inter-layer correlations are employed together to predict candidate macro-block modes and their encoding orders to exclude low possible modes, and then three new early termination methods are proposed.
- Human Action Recognition for Home Sensor Network4
This paper presents a survey on video-based human action recognition and addresses this problem from two major aspects: action representation and recognition approach.
- Reducing the Motion-Compensated Temporal Interpolation Noise of DVC Side Information by ODWT4
A novel MCTI-ODWT algorithm to generate better side information frame is introduced, which utilizes overcomplete discrete wavelet transform (ODWT) not only to reduce the noise caused by dense motion field error, but also to avoid the noise resulted from the high frequency components to the inaccurate motion vectors.
- 4
- 4
- 4
- 3
- 3
- MA-NeRF: Motion-Assisted Neural Radiance Fields for Face Synthesis from Sparse Images3
This work improves from NeRF and proposes a novel framework that, by leveraging the parametric 3DMM models, can reconstruct a high-fidelity drivable face avatar and successfully handle the unseen expressions.
- 3
- 3
- 3
- Federated Knowledge Transfer for Heterogeneous Visual Models3
FedDFA uses deep-layer knowledge distillation to align the feature representation and solve the knowledge transfer problem of heterogeneous models and facilitates FL with wide applicability for various models and better generalization performance than the state-of-the-art methods.
- Logic Enhanced Commonsense Inference with Chain Transformer3
The proposed Chain Transformer is a logic enhanced commonsense inference model that combines both direct and indirect inferences to construct a logical chain so as to reason in a more logically consistent way.
- Bridge the Gap: High-level Semantic Planning for Image Captioning3
A high-level semantic planning (HSP) mechanism that incorporates both a semantic reconstruction and an explicit order planning is proposed that outperforms previous methods and achieves the state-of-the-art performance on MS COCO.
- 3
- 3
- Improving Object Detection with Convolutional Neural Network via Iterative Mechanism3
This paper trains an iterative version of Faster RCNN to show the application of the iterative mechanism in improving the localization accuracy and presents a prototype CNN model that iteratively searches for objects on a very simple dataset to generate proposals.
- Learning Boltzmann machine with EM-like method3
It is proved that EM-like method will be the exactly same with contrastive divergence in restricted Boltzmann machine if the M-step of this method adopts special approximation.
- 3
- 3
- 3
- 3
- 3
- 3
- 2
- 2
- 2
- Bayesian Unsupervised Disentanglement of Anatomy and Geometry for Deep Groupwise Image Registration2
A novel hierarchical variational auto-encoding architecture is proposed to realise the inference procedure of the latent variables of the latent variables of groupwise image registration, where the registration parameters can be explicitly estimated in a mathematically interpretable fashion.
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- FPGA-based Acceleration System for Visual Tracking2
A real-time visual tracking algorithm based on DSST(Discriminative Scale Space Tracking) approach is proposed and a hardware system based on this algorithm is implemented on Xilinx XC7K325T FPGA platform.
- 2
- Conditional Kronecker Batch Normalization for Compositional Reasoning.2
A novel CBN method using the Kronecker transformation, termed as Conditional Kronecker Batch Normalization (CKBN), which outperforms previous CBN methods in compositional reasoning, counting as well as the convergence speed on CLEVR dataset.
- 2
- Enhanced Joint Trilateral Up-sampling for Super-Resolution2
This work uses low resolution depth image, refined high resolution color image and generated HR depth image to conduct iterative joint trilateral up-sampling and proposes a new depth image super-resolution method based on the assumption that LR image is a counterpart of HR image with missing pixels.
- 2
- Object Detection Based on Scene Understanding and Enhanced Proposals2
This paper studies the role of scene understanding in object detection by proposing a two-stream pipeline model called Scene-Object Network for Detection SOND, based on scene-like info and object detection method, which are separately implemented by two deep ConvNets and merged at the end of the pipeline.
- 2
- Fast and Robust Edge-Guided Exemplar-Based Image Inpainting2
The results show the proposed method can not only greatly reduce the computing time of exemplar-based methods, but also behave better on visual plausibility and continuity.
- 2
- 2
- Feature Extraction Methods for Handwritten Character Recognition Based on Acceleration.2
In the coding and decoding of Reed-Solomon codes formed of symbols larger than information symbols, redundant circuitry is eliminated, error detection and correction are performed using a simple construction, and the reliability of error detectionand correction is improved by processing only data of the same size as the information symbols.
- 2
- 2
- A Time-Event Mapping Method Based Transformation2
The experiment indicates that the transformation based error driven learning is a good patch for based rule method for the determining of mapping relation between event and its temporal expression.
- Using Traffic Stream Characteristics of Conflict Point to Determin Signal Time of Signalized Intersection2
Based on the gap acceptance theory, the formula of calculating signal time is deduced by using running modes of the stream at the conflict point, which is more efficient in real time control.
- 1
- One-Step Image Reconstruction for Cine MRI with a Quadratic Constraint1
This work derives a non-iterative algorithm to minimize a constrained objective function, which uses the measurements at other time frames to regularize the image reconstruction in cine MRI.
- 1
- 3D Isotropic black-blood cine MRI of intracranial arteries1
Comparison of the black-blood with 2D bright-blood PC-MRA and assessment of luminal boundary changes across the cardiac cycle suggest that pulsation of large intracranial arteries can be detected and measured by black-blood cine MRI.
- 1
- 1
- 1
- 1
- StrokeNet: Stroke Assisted and Hierarchical Graph Reasoning Networks1
SAPN is introduced, which performs hierarchical representation learning of text areas, effectively capturing extremely small or low-resolution texts and inferring structural relations between the hierarchical representations of each text area in the graph-based network.
- 1
- 1
- 1
- 1
- 1
- Multi-Scale Visual Semantics Aggregation with Self-Attention for End-to-End Image-Text Matching1
Self-Attention Visual-semantic Embeddings (SAVE) is proposed, which aggregates instance-level semantics from all potential positions of the image in an end-to-end manner and demonstrates both effectiveness and applicability of the method with favorably competitive performance as the state-of-the-art approaches.
- 1
- 1
- 1
- Research on Effective Evaluation of POL Support System Based on the Internet of Things and Cloud Theory1
The effectiveness evaluation model of POL support system is established based on cloud theory and it can be learned that the POL support efficiency based on the internet of things is excellent.
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- A Classified P2P Overlay Scheme Using SVC for Video Streaming1
The characteristics of Scalable Video Coding and Peer-to-Peer networks are leveraged in order to present an efficient and adaptive mechanism to integrate the P2P system and SVC together for video media streaming.
- Broadcast Encryption-Based P2P DRM without Central License Server1
This paper describes a basic implementation of broadcast encryption-based P2P DRM system, presents a general scheme with focus on how to decentralize the license agency to reserve the flexibility of P 2P network, and analyzes its security, efficiency and tradeoff between the two system parameters.
- Trace and revoke systems with short ciphertexts1
This paper presents a new fully collusion resistant trace and revoke system which has only constant size ciphertexts and is CPA secure against static adversaries with public broadcast key and secret tracing key.
- 1
- An algorithm for motion and change detection in image sequences based on chaos and information theory1
This paper analyzes image sequences using phase plots generated from sequential image frames and demonstrates that the changes in pixel amplitudes due to the motion of objects in an image sequence result in phase space behaviour resembling a chaotic signal.
- A Novel Hierarchical Mode Selection Algorithm for P-Slices in H.264/AVC1
A novel hierarchical mode selection algorithm is proposed to reduce the complexity of the mode selection process for P-Slices in H.264/AVC with a flexible early termination strategy that can efficiently skip the unnecessary mode checking while achieving a high prediction accuracy.
- 1
- 1
- Scalable Protection for MPEG-41
Two novel encryption algorithms for MPEG-4 FGS are proposed that are fast, error-resilient, and have negligible compression overhead and can be applied to other scalable multimedia formats.
- Constraint-preserving architecture transformations: A graph rewriting approach1
The authors characterize architecture transformations using graph rewriting rules, where architectures are represented in graph notations, to construct the sufficient and necessary condition for a transformation to preserve a constraint.
- –
- –
- –
- –
- –
- –
- –
- Voxel-Wise Quantification and Multi-planar Visualization of Plaque Contrast Enhancement in Intracranial Vessel Wall MRI–
The proposed CE map demonstrated good accuracy compared to expert review and high scan-rescan reproducibility, enabling comprehensive assessment of CE presence, intensity, and volume for evaluating plaque vulnerability in intracranial atherosclerotic disease.
- –
- MIGA: Mutual Information-Guided Attack on Denoising Models for Semantic Manipulation–
This paper proposes Mutual Information-Guided Attack (MIGA), the first method designed to directly attack deep denoising models by strategically disrupting their ability to preserve semantic content via adversarial perturbations, demonstrating its consistent effectiveness in disrupting semantic fidelity.
- –
- –
- VesselVoyager: An Interactive 3D Intracranial Vessel Tracing Software–
Voyager is a tool that facilitates detailed 3D mapping and analysis of cerebral vasculature, filling a critical gap in neurovascular diagnostic technology and enhances centerline tracing precision and triples processing efficiency, thus lightening the workload for clinicians and researchers.
- –
- Improved Simultaneous Intracranial Vessel Wall Imaging and dynamic MRA Using Low-rank Reconstruction–
This study aims to develop a novel low-rank-based image reconstruction method to reconstruct VW and dMRA from highly undersampled 4D MRI, and achieves high quality VW images with sharper wall delineation, while hemodynamic information from dMRA is preserved.
- –
- Effi–
This research proposes a framework for effective model adaptation using meta-models plus latent variables, which is inferred by the transformer-based system recognition module trained in an unsupervised fashion, and achieves a model with reduced prediction error and facilitating efficient adaptation when compared to prior dynamic model generalization methods and OMRL algorithms.
- Construction of a contrast enhancement map of intracranial vessel wall MRI with 3D volume rendering–
A CE map with 3D volume rendering, extended from a previously developed intracranial atherosclerosis analysis pipeline MOCHA, demonstrates CE’s spatial location relative to the vessel and provides reliable quantification of its intensity.
- –
- Towards Arbitrary Text-driven Image Manipulation via Space Alignment–
This work proposes a new Text-driven image Manipulation framework via Space Alignment (TMSA), which provides the user with an interface to control the attributes of a given image according to text input and get the result in real time.
- –
- Improving the Latent Space of Image Style Transfer–
Two contrastive training schemes are proposed to solve issues in the latent space used by style transfer, where the style contrastive loss pulls the stylized result closer to the same visual style image and pushes it away from the content image.
- –
- VQMG: Hierarchical Vector Quantised and Multi-hops Graph Reasoning for Explicit Representation Learning–
VQMG, a novel and unified framework for multi-hops relational reasoning and explicit representation learning is proposed by introducing Multi-hops Graph Convolution Networks (MGCN), complicated relations from hierarchical latent space are effectively captured by Inner Graph, while the fitting of autoregressive prior are performed coherently by Outer Graph to promote the performance.
- –
- –
- An Inverse Mapping with Manifold Alignment for Zero-Shot Learning–
This work proposes the inverse mapping, which predicts the visual feature prototype of each unseen class via the semantic vector for image classification, and proposes a heuristic algorithm to select a high density set from data of each seen class.
- Saliency Detection based on Depth and Sparse Features–
This paper modified the region-based Human Visual System (HVS) model by import two features, sparse feature and depth feature, and shows that the modified method works better and can obtain a more accurate result.
- –
- Human Parsing via Shape Boltzmann Machine Networks–
An approach is presented that exploits Shape Boltzmann Machine networks to improve the accuracy of human body parsing, and the proposed Curve Correction method refines the final segmentation results.
- Improved TZsearch Fast Search Algorithm for JMVC–
On the premise of guaranteeing the quality of reconstructed video and coding bit-rate, the experiment result shows that in the test of sequences with different features the improved algorithm takes only 40–60 % total encoding time of the former fast algorithm.
- –
- A Distributed Video Coding Algorithm Based on Weighted Motion Estimation–
A way to improve the prediction accuracy, during the process of generating predictive side information, makes full use of the distributed structural characteristics, and improves the RD performance of DVC by using bidirectional motion estimation and the weighted average methods is proposed.
- –
- An early Mode Decision Algorithm for the Enhancement Layers in H.264 Scalable Video Coding–
A early mode decision algorithm is proposed to reduce the complexity of the mode selection process for enhancement layers in H.264 Scalable Video Coding and results show that 30% encoding time can be saved with a negligible loss in BDSNR, and BDBR can be significantly reduced.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Novel Early Mode Decision Algorithm for Enhancement Layers in H.264 Scalable Video Coding–
A novel early mode decision algorithm is proposed to reduce the complexity of the mode selection process for enhancement layers in H.264 Scalable Video Coding and results show that 10%~40% encoding time can be saved.
- –
- ANALYSES ON REGIONAL CULTIVATED LAND CHANGEBASED ON QUANTITATIVE METHOD–
This paper uses the cultivated land areas and social economic indicators of reservoir area of Three Gorges in 1990-2004, takes the statistic analyses and example research in order to analyze the process of cultivated land, get the driving forces of cultivatedLand change, and find the new methods to stimulate and forecast the cultivatedLand areas in the future.
- –
- ANALYSES ON REGIONAL CULTIVATED LAND CHANGE BASED ON QUANTITATIVE METHOD Reservoir Area of Three Gorges as Example–
This paper uses the cultivated land areas and social economic indicators of reservoir area of Three Gorges in 1990-2004, takes the statistic analyses and example research in order to analyze the process of cultivated land, get the driving forces of cultivatedLand change, and find the new methods to stimulate and forecast the cultivatedLand areas in the future.
- The Design and Implementation of High-level Secure Operating System SECIMOS by the Way of Improvment/Enhancement–
The system framework and key techniques, including access control, accountability and assurance methods, are given and it indicates new suitable approach to develop the highest-level secure operating system should be researched further.
- –
- –
- –
- Innovative hyperchaotic encryption algorithm for compressed video–
The innovative hyperchaotic cryprography proves robust security by cryptoanalysis, shows good real-time performance and flexible implement capability through the arithmetic evaluating and test and implements encryption and identity authentification on control parameters at initialization phase.
- A research on the modified design of Post Codes–
Three modified designs of Post Codes together with the fundamentals of modification are brought forward and the analysis of availability and venture of each plan is beneficial to decision making.
- Detecting Objects In MR Image Sequences Based On Minimal Path Active Contour Model–
This research presents a novel and scalable approaches that allow for real-time, 3D image analysis of the dynamic response of the immune system to treat central giant cell granuloma.
- Chinese Grammar Rules Learning Based on Syntactic and Semantic Knowledge–
Experimental results demonstrate that Chinese grammar rules could be acquired automatically and the complexity of Chinese parsing could be reduced also in the system.
- –
- –
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.