Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks88 from public data
- Not only Look, But Also Listen: Learning Multimodal Violence Detection Under Weak Supervision616
A neural network containing three parallel branches to capture different relations among video snippets and integrate features, where holistic branch captures long-range dependencies using similarity prior, localized branch captures local positional relation using proximity prior, and score branch dynamically captures the closeness of predicted score.
- HVI: A New Color Space for Low-light Image Enhancement244
A new color space for LLIE is proposed, namely Horizontal/Vertical-Intensity (HVI), defined by polarized HS maps and learnable intensity, defined by polarized HS maps and learnable intensity.
- 226
- 224
- 221
- 198
- ASCU-Net: Attention Gate, Spatial and Channel Attention U-Net for Skin Lesion Segmentation154
This paper proposes an extended version of U-Net for the segmentation of skin lesions using the concept of the triple attention mechanism, and demonstrates the effectiveness to the presence of irregular borders, lesion and skin smooth transitions, noise, and artifacts.
- Video Event Restoration Based on Keyframes for Video Anomaly Detection128
A novel U-shaped Swin Transformer Network with Dual Skip Connections (USTN-DSC) for video event restoration, where a cross-attention and a temporal upsampling residual skip connection are introduced to further assist in restoring complex static and dynamic motion object features in the video.
- Open-Vocabulary Video Anomaly Detection123
This paper proposes a model that decouples OVVAD into two mutually comple-mentary tasks - class-agnostic detection and class-specific classification - and jointly optimizes both tasks, and designs a semantic knowledge injection module and novel anomaly synthesis module to generate pseudo unseen anomaly videos with the help of large vision generation models for the classification task.
- 119
- Fast sparse coding networks for anomaly detection in videos105
A Fast Sparse Coding Network (FSCN) based on High-level Features is proposed, which is hundreds of or even thousands of times faster at the test stage and reaches the state-of-the-art level.
- Weakly Supervised Video Anomaly Detection and Localization with Spatio-Temporal Prompts96
A novel method called STPrompt that learns spatio-temporal prompt embeddings for weakly supervised video anomaly detection and localization (WSVADL) based on pre-trained vision-language models (VLMs) achieves state-of-the-art performance on three public benchmarks for the WSVADL task.
- 86
- 82
- HANet79
The proposed Hierarchical Alignment Network (HANet) to align different level representations for video-text matching outperforms other state-of-the-art methods, which demonstrates the effectiveness of hierarchical representation and alignment.
- Toward Video Anomaly Retrieval From Video Anomaly Detection: New Benchmarks and Model74
A novel task called Video Anomaly Retrieval (VAR), which aims to pragmatically retrieve relevant anomalous videos by cross-modalities, e.g., language descriptions and synchronous audios, and introduces an efficient pretext task to enhance semantic associations between video-text fine-grained representations.
- Exploiting foreground and background separation for prohibited item detection in overlapping X-Ray images58
A foreground and background separation X-ray prohibited item detection framework, which separates prohibited items from other items to exclude irrelevant information and significantly outperforms state-of-the-art solutions.
- 47
- Bracketing Image Restoration and Enhancement with High-Low Frequency Decomposition41
Inspired by the notion that high/low frequency information is applicable to different degradations, HLNet is introduced, a Bracketing Image Restoration and Enhancement method based on high-low frequency decomposition, thus achieving higher-quality image restoration.
- Automatic Modulation Classification Based on Deep Learning for Software-Defined Radio41
An improved deep neural architecture for implementing radio signal identification tasks, which is an important facet of constructing the spectrum-sensing capability required by software-defined radio, based on the Inception-ResNet network.
- CRNet: A Detail-Preserving Network for Unified Image Restoration and Enhancement Task40
The Composite Refinement Network (CRNet) is proposed, which can perform unified image restoration and enhancement on multiple exposure images using multiple exposure images, and explicitly separates and strengthens high and low-frequency information through pooling layers, using specially designed Multi-Branch Blocks.
- 39
- Bidirectional Retrospective Generation Adversarial Network for Anomaly Detection in Videos37
A novel future frame prediction model based on a bidirectional retrospective generation adversarial network (BR-GAN) for anomaly detection that outperforms most existing state-of-the-art methods and a sequence discriminator composed of a 3D convolutional neural network to capture the long-term temporal relationships between frame sequences composed of predicted frames and input frames.
- Automatic stent reconstruction in optical coherence tomography based on a deep convolutional model36
The deep-convolutional model can accurately detect stent struts in IVOCT images, thus enabling the fully-automatic quantification of stent parameters in an extremely short time, and might facilitate the application of quantitative IVO CT analysis in real-world clinical scenarios.
- NTIRE 2025 Challenge on Efficient Burst HDR and Restoration: Datasets, Methods, and Results30
A comprehensive overview of the NTIRE 2025 Efficient Burst HDR and Restoration Challenge is provided, the proposed solutions are compared, and it serves as a valuable reference for researchers and practitioners in efficient burst HDR and restoration.
- NTIRE 2025 Challenge on Low Light Image Enhancement: Methods and Results30
This paper thoroughly evaluates the state-of-the-art advancements in LLIE, showcasing the significant progress.
- Learning Optimal Individualized Treatment Rules from Electronic Health Record Data26
This work proposes a statistical method to estimate the optimal individualized treatment rules (ITRs) that are tailored according to subject-specific features using electronic health records (EHR) data and creates a broad range of feature variables that reflect both patient health status and healthcare data collection process.
- Boosting HDR Image Reconstruction via Semantic Knowledge Transfer25
This framework introduces the Semantic Priors Guided Reconstruction Model (SPGRM), which leverages SDR image semantic knowledge to address ill-posed problems in the initial HDR reconstruction results, and leverages a self-distillation mechanism that constrains the color and content information with semantic knowledge, aligning the external outputs between the baseline and SPGRM.
- 18
- The Collaborative System Workflow Management of Industrial Design Based on Hierarchical Colored Petri-Net16
A hierarchical colored Petri net (HCPN) is proposed for modeling the industrial design collaborative system workflow that provides the necessary information and knowledge resources for the design reuse, flexibility management, and intelligent manufacturing of subsequent products.
- 15
- IIOT-Based Intelligent Control and Management System for Motorcycle Endurance Test15
A four-layer IIOT architecture, that is perception executive layer (data acquisition), cognitive layer (business logic, identification, classification), network layer (communication layer), and control layer (external participation), is proposed aiming to realize the intelligent management and remote control of motorcycle ’s endurance test.
- 14
- EOGT: Video Anomaly Detection with Enhanced Object Information and Global Temporal Dependency13
A video anomaly detection framework with Enhanced Object Information and Global Temporal Dependencies (EOGT) and the main novelties are a Local Object Anomaly Stream (LOAS) and a Global Temporal Strengthening Stream (GTSS) with video-level temporal dependencies, which leverages video-level temporal dependencies to identify long-term and video-specific anomalies effectively.
- Fault Identification Ability of a Robust Deeply Integrated GNSS/INS System Assisted by Convolutional Neural Networks13
A novel robust algorithm based on convolutional neural network (CNN) is proposed which can successfully suppress the fault propagation and the accuracy of fault identification is increased, and the robustness of deep GNSS/INS integration is also improved.
- 12
- SlowFastVAD: Video Anomaly Detection via Integrating Simple Detector and RAG-Enhanced Vision-Language Model10
Inspired by dual complementary pathways in human visual perception, this work proposes SlowFastVAD, a hybrid framework that integrates a fast anomaly detector with a slow anomaly detector (namely a retrieval augmented generation (RAG) enhanced VLM), to address limitations of vision-language models (VLMs).
- 10
- 10
- Filter-GAN: Imbalanced Malicious Traffic Classification Based on Generative Adversarial Networks with Filter10
A feature image representation method and Adversarial Generative Network with Filter (Filter-GAN) are proposed to solve the problems of malicious traffic collection and labeling and significantly improved the recognition accuracy of small-sample datasets.
- 9
- Infrared Target Detection Based on Image Enhancement and an Improved Feature Extraction Network8
Real-world validation experiments confirm the effectiveness and efficiency of the proposed approach for practical UAV detection applications, and maintain a lightweight profile with only 6.2M parameters and enables real-time inference, which is crucial for practical deployment.
- 8
- 8
- Feature Selection for Facebook Feed Ranking System via a Group-Sparsity-Regularized Training Algorithm8
A novel neural-network-suitable feature selection algorithm, which selects important features from the input layer during training by injecting group-sparsity regularization into the (stochastic) training algorithm.
- 7
- FusionNet: Multi-Model Linear Fusion Framework for Low-Light Image Enhancement7
FusionNet is introduced, a novel multi-model linear fusion framework that operates in parallel to effectively capture global and local features across diverse color spaces by incorporating a linear fusion strategy underpinned by Hilbert space theoretical guarantees.
- 6
- 6
- 6
- 5
- Research on WSN reliable ranging and positioning algorithm for forest environment5
This paper presents a reliable WSN distance measurement-positioning algorithm for forest environments that derives the reference points of the logarithmic distance path loss model and path loss index and locates target nodes using anchor nodes in different regions.
- Stress Wave Analysis Based Prognostic Health Management5
The stress wave sensor detect and process the electronic signal of friction, mechanical shock and dynamic load on equipment moving parts, the stress wave analysis are fulfilled by using the time domain and frequency domain feature extraction software, Polynomial neural network (PNN) and data fusion technology.
- HVI-CIDNet+: Beyond Extreme Darkness for Low-Light Image Enhancement4
This work proposes HVI-CIDNet+, a Color and Intensity Decoupling Network designed to restore degraded content and mitigate color distortion in extremely dark regions, and introduces a Region Refinement Block that combines convolution for information-rich regions with attention for information-scarce regions, enabling accurate brightness adjustment.
- Human-Centric Behavior Description in Videos: New Benchmark and Model4
A human-centric video surveillance captioning dataset is constructed, which provides detailed descriptions of the dynamic behaviors of 7,820 individuals, and a novel video captioning approach is proposed that can describe individual behavior in detail on a person-level basis, achieving state-of-the-art results.
- Remote Sensing Image Fusion Based on Morphological Convolutional Neural Networks with Information Entropy for Optimal Scale4
Experimental results show that the remote sensing image fusion method proposed can effectively retain the information of the original image, improve the spatial resolution and spectral fidelity, and provide a new idea for image fusion from the perspective of multi-morphological deep learning.
- Design of Equipment Online Monitoring and Fault Diagnosis Platform Based on IOT4
The platform of this paper presented is mainly based on stress wave condition monitoring technology, which combined with process monitoring to form an online monitoring data set of the equipment and combines principal component analysis and the BP neural network to predict the equipment’s health trend and fault type.
- Double δ-LBP: A Novel Feature Extraction Method for Facial Expression Recognition4
Two δ-LBPs are employed to represent details and the contour of faces separately, which take different kinds of information of facial expression into account, and show that Dδ-LBP outperforms the original LBP method.
- 2
- 2
- 2
- 2
- Transformer multimodal data and knowledge encoding method based on attention model2
Multimodal data knowledge encoding technology is applied to transformer fault diagnosis and shows that the fault diagnosis accuracy can be significantly improved after fusion and encoding of multimodal data.
- 2
- Multi-Region Two-Stream Deep Architecture for Visual Power Monitoring Systems2
A novel multi-region two-stream deep architecture to improve judging abnormalities that incorporates two- stream scheme and multi- region strategy to identify relevant information and explore hidden details, and provides an end-to-end learning procedure.
- 2
- 2
- Multi-disciplinary Design Optimization of Air-Breathing Hypersonic Vehicle Using Pareto Games and Evolutionary Algorithms2
A Pareto-EAs methodology is developed, which permits a quick performance evaluation of an idealized, integrated SCRAMJET vehicle for preliminary design analysis and is well suited for its implementation on HPCs for increasing its efficiency.
- 2
- 1
- 1
- 1
- 1
- A crack detection network based on deformable convolution and test time augmentation1
An improved crack detection network, which can provide convenience in pavement maintenance and reduce the burden of labor costs, is proposed based on YOLOv5-l, which has a balance of accuracy and speed.
- 1
- Texture and Wavelet based PCANet (TW-PCANet) for Infants' Facial Expression Recognition1
The texture and wavelet based PCANet (TW-PCANet) is proposed to fully extract the features of infants' facial expressions, and the experimental results show that the proposed method can achieve higher recognition rate obtained by PCANnet, LBP and Gabor wavelet.
- An Online Kernel Learning Algorithm based on Orthogonal Matching Pursuit1
The orthogonal matching pursuit is employed not only to guide the online learning algorithm to estimate the target function but also to keep control of the sparsity of the solution.
- –
- –
- STPrompt \(\boldsymbol{++}\) : Prompting Vision-Language Models for Weakly Supervised Video Anomaly Detection and Fine-Grained Localization–
This work systematically leverage preliminary coarse localization regions derived from anomaly scores as spatial priors, together with contextual frames around keyframes, zoomed-in views of suspected anomalous regions, and refined textual descriptions of anomalies to guide the VLMs toward deep semantic comprehension of video anomalies, enabling accurate pixel-level spatial localization.
- –
- –
- –
- –
- MGCAN: A Novel Method for Visual Power Monitoring Systems–
A novel deep architecture to improve the assessment of abnormalities by considering a novel mutual-guided convolutional attention network (MGCAN) and a multi-region scheme that provides an end-to-end and adaptive learning procedure relevant to quality perception.
- Research and Implementation of Violation Monitoring System for Electric Customer Service Hall–
An electric customer service hall monitoring system based on Improved Tiny-yolo v3 is developed for common violation in the services of electric customer hall: mobile phone playing behavior, smoking behavior, sleeping behavior, manicuring behavior and fighting behavior.
- –
- –
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 2026-10-15. 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.