Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks244 from public data
- Transformer Tracking1,440
This work presents a novel attention-based feature fusion network, which effectively combines the template and search region features solely using attention and presents a Transformer tracking method based on the Siamese-like feature extraction backbone, the designed attention- based fusion mechanism, and the classification and regression head.
- 1,288
- Learning Spatio-Temporal Transformer for Visual Tracking1,225
A new tracking architecture with an encoder-decoder transformer as the key component, which models the global spatio-temporal feature dependencies between target objects and search regions, while the decoder learns a query embedding to predict the spatial positions of the target objects.
- 1,147
- Deep visual tracking: Review and experimental comparison517
The background of deep visual tracking is introduced, including the fundamental concepts of visual tracking and related deep learning algorithms, and the existing deep-learning-based trackers are categorize into three classes according to network structure, network function and network training.
- 496
- Learning Uncertain Convolutional Features for Accurate Saliency Detection480
A novel deep fully convolutional network model for accurate salient object detection and an effective hybrid upsampling method to reduce the checkerboard artifacts of deconvolution operators in the authors' decoder network are proposed.
- 462
- 434
- The Eighth Visual Object Tracking VOT2020 Challenge Results350
A significant novelty is introduction of a new VOT short-term tracking evaluation methodology, and introduction of segmentation ground truth in the VOT-ST2020 challenge – bounding boxes will no longer be used in theVDT challenges.
- GradNet: Gradient-Guided Network for Visual Object Tracking334
A novel gradient-guided network to exploit the discriminative information in gradients and update the template in the siamese network through feed-forward and backward operations and a template generalization training method is proposed to better use gradient information and avoid overfitting.
- Boosting Continual Learning of Vision-Language Models via Mixture-of-Experts Adapters311
This work presents a parameter-efficient continual learning framework to alleviate long-term forgetting in incremental learning with vision-language models, and introduces a Distribution Discriminative Auto-Selector (DDAS) that automatically routes in-distribution and out-of-distribution inputs to the MoE Adapter and the original CLIP, respectively.
- Vision-Based Anti-UAV Detection and Tracking307
A clear and simple tracking algorithm combined with detection that inherits the detector’s high precision is proposed, thus providing a new attempt at UAV tracking using the proposed DUT Anti-UAV dataset.
- Visible-Thermal UAV Tracking: A Large-Scale Benchmark and New Baseline298
A large-scale benchmark with high diversity for visible-thermal UAV tracking (VTUAV), including 500 sequences with 1.7 million high-resolution (1920* 1080 pixels) frame pairs is constructed, and a coarse-to-fine attribute annotation is provided, where frame-level attributes are provided to exploit the potential of challenge-specific trackers.
- High-Performance Long-Term Tracking With Meta-Updater289
This work proposes a novel offline-trained Meta-Updater that can effectively integrate geometric, discriminative, and appearance cues in a sequential manner, and then mine the sequential information with a designed cascaded LSTM module.
- 273
- Structured Siamese Network for Real-Time Visual Tracking266
A local structure learning method, which simultaneously considers the local patterns of the target and their structural relationships for more accurate target tracking and can be formulated as the inference process of a conditional random field and implemented by differentiable operations, allowing the entire model to be trained in an end-to-end manner.
- LightTrack: Finding Lightweight Neural Networks for Object Tracking via One-Shot Architecture Search265
This work presents LightTrack, which uses neural architecture search (NAS) to design more lightweight and efficient object trackers that achieve superior performance compared to handcrafted SOTA trackers, such as SiamRPN++ and Ocean, while using much fewer model Flops and parameters.
- Alpha-Refine: Boosting Tracking Performance by Precise Bounding Box Estimation232
This work proposes a novel, flexible, and accurate refinement module called Alpha-Refine (AR), which can significantly improve the base trackers’ box estimation quality and leads to a series of strengthened trackers, among which the ARSiamRPN (AR strengthened Siam RPNpp) and the ARDiMP50 ( AR strengthened DiMP50) achieve good efficiency-precision trade-off.
- Learning Spatial-Aware Regressions for Visual Tracking221
This paper proposes a kernelized ridge regression model wherein the kernel value is defined as the weighted sum of similarity scores of all pairs of patches between two samples, and shows that this model can be formulated as a neural network and thus can be efficiently solved.
- Jointly Modeling Motion and Appearance Cues for Robust RGB-T Tracking206
This study develops a novel late fusion method to infer the fusion weight maps of both RGB and thermal (T) modalities and proposes a tracker switcher to switch the appearance and motion trackers flexibly.
- ‘Skimming-Perusal’ Tracking: A Framework for Real-Time and Robust Long-Term Tracking205
This work presents a novel robust and real-time long-term tracking framework based on the proposed skimming and perusal modules, designed to efficiently choose the most possible regions from a large number of sliding windows for image-wide global search.
- Correlation Tracking via Joint Discrimination and Reliability Learning189
This work proposes a novel CF-based optimization problem to jointly model the discrimination and reliability information and introduces a local response consistency regular term to emphasize equal contributions of different regions and avoid the tracker being dominated by unreliable regions.
- Learning Adaptive Attribute-Driven Representation for Real-Time RGB-T Tracking186
This work observing that the implicit attribute information can boost the model discriminability, and proposing a novel attribute-driven representation network to improve the RGB-T tracking performance, which achieves very competitive results with a real-time tracking speed.
- 157
- 149
- 148
- 146
- Real-Time ‘Actor-Critic’ Tracking139
This work modifications the original deep deterministic policy gradient algorithm to effectively train the ‘Actor-Critic’ model for the tracking task and demonstrates that the proposed tracker performs favorably against many state-of-the-art methods, with real-time performance.
- 130
- 129
- The Ninth Visual Object Tracking VOT2021 Challenge Results127
The Visual Object Tracking challenge VOT2021 is the ninth annual tracker benchmarking activity organized by the VOT initiative; results of 71 trackers are presented; many are state-of-the-art trackers published at major computer vision conferences or in journals in recent years.
- 118
- 118
- Non-rigid object tracking via deep multi-scale spatial-temporal discriminative saliency maps109
The proposed algorithm achieves competitive performance in comparison with state-of-the-art methods for both saliency detection and visual tracking, especially outperforming other related trackers on the non-rigid object tracking datasets.
- 106
- ROI Pooled Correlation Filters for Visual Tracking101
A novel ROI pooled correlation filter (RPCF) algorithm for robust visual tracking is proposed and mathematical derivations show that the ROI-based pooling can be equivalently achieved by enforcing additional constraints on the learned filter weights, which makes theROI- based pooling feasible on the virtual circular samples.
- 87
- 86
- 84
- 82
- High-Performance Transformer Tracking76
To determine whether a better feature fusion method exists than correlation, a novel attention-based feature fusion network, inspired by the transformer, is presented and effectively combines the template and search region features using attention mechanism.
- Multi attention module for visual tracking75
This work proposes a new visual tracking algorithm leveraging multi-level visual attention to take full use of the information during tracking, and implements the attention network with long short term memory (LSTM) units, which are capable of capturing the historical context information to perform more reliable inference at the current time step.
- 73
- Latent Distribution Preserving Deep Subspace Clustering73
A novel deep subspace clustering method based on a latent distribution-preserving autoencoder is proposed, which introduces a distribution consistency loss to guide the learning of distribution- Preserving latent representation, and consequently enables strong capacity of characterizing the real-world data for sub space clustering.
- 67
- Kernel collaborative face recognition66
A general formulation known as kernel collaborative representation is presented to encompass several effective representation-based classifiers within a unified framework and an effective algorithm with squared L2-regularization is developed.
- Deep mutual learning for visual object tracking65
Extensive experiments on the OTB2013, OTB2015, VOT2017 and LaSOT benchmarks demonstrate that the tracking performance can be improved effectively by using the proposed mutual-learning-based training methodology.
- Deep-Sea Organisms Tracking Using Dehazing and Deep Learning63
This paper develops a method for underwater real-time recognition and tracking of multi-objects, which is called “You Only Look Once: YOLO”, and provides a very fast and accurate tracker.
- 62
- 60
- 59
- Spectral-spatial K-Nearest Neighbor approach for hyperspectral image classification58
This paper proposes a spectral-spatial KNN (SSKNN) method to deal with the HSI classification problem, which effectively exploits the distances all neighboring pixels of a given test pixel and training samples.
- SUTrack: Towards Simple and Unified Single Object Tracking53
A simple yet unified single object tracking framework that consolidates five SOT tasks into a unified model trained in a single session, and provides a range of models catering edge devices as well as high-performance GPUs, striking a good trade-off between speed and accuracy.
- Multi-modal visual tracking: Review and experimental comparison48
Different aspects of multi-modal tracking algorithms are summarized under a unified taxonomy, with specific focus on visible-depth (RGB-D) and visible-thermal (RGB-T) tracking.
- 47
- Two Stage Shot Boundary Detection via Feature Fusion and Spatial-Temporal Convolutional Neural Networks45
A two stage method for shot boundary detection (TSSBD) is presented which distinguishes abrupt shot by fusing color histogram and deep features, and locate gradual shot changes with C3D-based deep analysis.
- EEG-Based Detection of Epileptic Seizures Through the Use of a Directed Transfer Function Method43
The results indicated that the DTF-derived connectivity could characterize the dynamic causal interaction patterns between brain areas during seizure states, and the proposed method was suitable for the detection of epileptic seizures.
- 41
- Adaptive spatial pixel-level feature fusion network for multispectral pedestrian detection40
The extensive experiments on the KAIST dataset demonstrate the effectiveness of the proposed method for the fusion of visible and infrared image in multispectral pedestrian detection.
- MambaVT: Spatio-Temporal Contextual Modeling for Robust RGB-T Tracking37
This work innovative proposes a pure Mamba-based framework (MambaVT) to fully exploit spatio-temporal contextual modeling for robust visible-thermal tracking, and devise the long-range cross-frame integration component to globally adapt to target appearance variations.
- Exploring Dynamic Transformer for Efficient Object Tracking36
This article proposes DyTrack, a dynamic transformer framework for efficient tracking that automatically learns to configure proper reasoning routes for different inputs, thereby improving the utilization of the available computational budget and achieving higher performance at the same running speed.
- 35
- 35
- 34
- 34
- 30
- 30
- 30
- 29
- Learning 2D Invariant Affordance Knowledge for 3D Affordance Grounding28
This work introduces the Multi-Image Guided Invariant-Feature-Aware 3D Affordance Grounding (MIFAG) framework and proposes to learn generalizable invariant affordance knowledge from multiple human-object interaction images within the same affordance category.
- Learning Regression and Verification Networks for Robust Long-term Tracking28
A new visual tracking algorithm is proposed, which leverages the merits of both template matching approaches and classification models for long-term object detection and tracking and achieves state-of-the-art performance on the OxUvA long- term tracking dataset.
- 27
- 26
- 26
- 25
- Multi-feature tracking via adaptive weights25
This work exploits intensity, histogram of gradient (HOG) and color naming features to generate a set of confidence maps where the confidence value of each pixel indicates the probability that this pixel belongs to the tracked object.
- 25
- Aeroengine remaining useful life prediction via integrating enhanced inverted transformer and spatiotemporal graph learning24
Experimental evaluations on the C-MAPSS dataset demonstrate that FIT-GSTM achieves superior RUL prediction accuracy and generalization compared to existing methods, highlighting its potential for real-world deployment in aeroengine health management.
- Target attention deep neural network for infrared image enhancement24
An innovative target attention deep neural network (TADNN) is presented to realize a discriminative enhancement in an end-to-end manner and outperforms the competitive approaches on both of subjective visual effect and quantitative assessments.
- Long-term Visual Tracking: Review and Experimental Comparison23
This paper provides a thorough review of long-term tracking, summarizing long- term tracking algorithms from two perspectives: framework architectures and utilization of intermediate tracking results, and discusses the future prospects from multiple perspectives.
- Weighted Generalized Nearest Neighbor for Hyperspectral Image Classification23
An effective classification framework to classify a hyperspectral image (HSI), which consists of two fundamental components: weighted generalized nearest neighbor (WGNN) and label refinement, and a novel label refinement method to enhance label consistency in the classification process.
- 23
- 23
- 22
- Spatial-temporal initialization dilemma: towards realistic visual tracking21
This paper proposes a simple compensation framework that contains spatial-refine and temporal-chasing modules to mitigate performance degradation caused by the initialization dilemma and can be compatible with various popular trackers without retraining.
- 21
- VELIE: A Vehicle-Based Efficient Low-Light Image Enhancement Method for Intelligent Vehicles20
A deep learning enhancement network named Vehicle-based Efficient Low-light Image Enhancement (VELIE) achieves state-of-the-art performance on various driving datasets with a processing time of only 0.19 s, significantly enhancing high-dimensional environmental perception tasks in low-light conditions.
- 20
- Leveraging the Power of Data Augmentation for Transformer-based Tracking19
This paper explores the impact of general data augmentations on transformer-based trackers via systematic experiments, and proposes two data augmentation methods customized for tracking, including a token-level feature mixing augmentation strategy which enables the model against challenges like background interference.
- 19
- 19
- 17
- 16
- Using Weakly Supervised Deep Learning to Classify and Segment Single‐Molecule Break‐Junction Conductance Traces15
This work proposes a weakly supervised deep learning algorithm to classify and segment single-molecule break junction traces, a method that is mainly based on transfer learning with the pretrain-finetune technique and has achieved a better accuracy-to-manpower efficiency balance.
- Encoder combined video moving object detection14
Experiments on a number of video sequences demonstrate the effectiveness in detection accuracy and superior processing speed of the proposed low computation moving object detection method combined with video encoder.
- 14
- 13
- Balanced Gradient Penalty Improves Deep Long-Tailed Learning13
A universal Balanced Gradient Penalty (BGP) method, which can achieve the state-of-the-art results on various image datasets in the settings of different imbalance ratios, and does not need the detailed class priors and preserves privacy.
- Online Filtering Training Samples for Robust Visual Tracking12
An optimization module named MetricNet is proposed for online filtering training samples that applies a matching network containing the classification and distance branches, and uses multiple metric methods for different type samples to improve the tracking performance.
- 12
- 12
- 12
- SRRT: Exploring Search Region Regulation for Visual Object Tracking11
A novel tracking paradigm, called Search Region Regulation Tracking (SRRT) that applies a small eyereach when the target is captured and zooms out the search field when the target is about to be lost, which achieves evident improvements and competitive results with other state-of-the-art trackers on eight benchmarks.
- 11
- A Facial Expression Recognition System for Smart Learning Based on YOLO and Vision Transformer11
This paper proposes a facial expression recognition system for smart learning on classroom that uses YOLO to extract face images of multiple students from high-resolution video, then a self-attention based model named Vision Transformer is used to recognize facial expressions.
- 11
- 11
- Fast and effective color-based object tracking by boosted color distribution11
A novel BCD algorithm for color object tracking that achieves good results at a fast speed is presented and is better than or not worse than some state-of-the-art tracking algorithms on some public sequences.
- 10
- LiveScene: Language Embedding Interactive Radiance Fields for Physical Scene Rendering and Control10
This paper proposes LiveScene, a scene-level language-embedded interactive radiance field that efficiently reconstructs and controls multiple objects, and demonstrates significant superiority in novel view synthesis, interactive scene control, and language grounding performance through extensive experiments.
- 10
- Video Annotation for Visual Tracking via Selection and Refinement10
A new framework to facilitate bounding box annotations for video sequences is presented, which investigates a selection-and-refinement strategy to automatically improve the preliminary annotations generated by tracking algorithms.
- Fast 3D Semantic Mapping in Road Scenes10
This work proposes a fast 3D semantic mapping system based on the monocular vision by fusion of localization, mapping, and scene parsing that improves the accuracy of the3D semantic labeling in light of the spatial consistency on each step of 3D reconstruction.
- Visual-attention gabor filter based online multi-armored target tracking9
This work proposed a more efficient and effective spatial-temporal attention scheme to track multi-armored target in the ground battlefield by simulating the structure of the retina, a novel visual-attention Gabor filter branch is proposed to enhance detection.
- 9
- 8
- 8
- 8
- Online single target tracking in WAMI: benchmark and evaluation8
By combining the basic appearance model, background subtraction and high-order motion (HoM) affinity, a novel normalized cross correlation HoM (NCC- HoM) tracking algorithm for WAMI OSTT is developed.
- Fast Pedestrian Detection Based on Sliding Window Filtering8
This paper proposes an effective and efficient pedestrian detection method based on sliding windows with well-designed multi-scale scanning over candidate regions instead of whole frames that outperforms the state-of-the-art methods in both speed and accuracy of detection.
- Lane Graph Extraction from Aerial Imagery via Lane Segmentation Refinement with Diffusion Models7
The lane mask refinement approach enhances the quality of the extracted lane graph, yielding gains of approximately 1.5\% in GEO F1 and 3.5\% in TOPO F1 scores over the best-performing CNN-based method, and improvements over the best-performing diffusion-based method.
- Research on Image Recognition Algorithm Technology for Power Line detection7
Experimental results show that the algorithm can detect power lines in aerial images better, and the algorithm is not complicated, which is suitable for real-time detection of power lines.
- Visual tracking by dynamic matching-classification network switching7
This work presents a novel real-time robust tracking method to take full use of the benefits from both kinds of networks, and proposes a matching-classification network switching (MCS) framework to integrate the matching, classification, verification networks and conduct dynamic switching among them.
- 7
- 6
- 6
- 6
- 5
- Exploiting Lightweight Hierarchical ViT and Dynamic Framework for Efficient Visual Tracking5
The core innovation of HiT lies in its Bridge Module, which connects lightweight transformers to the tracking framework, enhancing feature representation quality, and a training-free acceleration method based on the dynamic routing architecture of DyHiT is introduced.
- Efficient Motion Prompt Learning for Robust Visual Tracking5
A lightweight and plug-and-play motion prompt tracking method that can be easily integrated into existing vision-based trackers to build a joint tracking framework leveraging both motion and vision cues, thereby achieving robust tracking through efficient prompt learning.
- Feature Dense Relevance Network for Single Image Dehazing5
A novel dehazing network defined by defining the Feature Dense Relevance module (FDR) and the Shallow Feature Mapping module (SFM) that not only maps the degraded textures into clear textures by distant dependence, but also reduces artifacts and color distortions effectively.
- Transmission Line Obstacle Detection Based on Structural Constraint and Feature Fusion5
The validity of the proposed obstacle detection method based on structural constraints and feature fusion is verified and shows the effectiveness of weighted fusion of global and local features.
- 5
- 5
- Feature-preserving color pencil drawings from photographs4
To mimic the tonal style of color pencil drawings, which are much lighter and have relatively lower saturation than photographs, a lightness enhancement mapping and a saturation reduction mapping are devised.
- Fluid-inspired field representation for risk assessment in road scenes4
A fluid-inspired model is proposed to estimate collision risk in road scenes using the public KITTI dataset; this model can provide predictions in the cases of misdetection and tracking error caused by occlusion and proves a promising approach for collision risk assessment.
- 4
- Motion-guided Visual Tracking3
This study decouple motion into two types: camera motion and object motion, and proposes a novel self-supervised camera motion verifier to validate its confidence and effectiveness, and demonstrates the effectiveness and generalization ability of the module.
- 3
- 3
- 3
- 3
- 3
- Image segmentation based on gray equalization and improved genetic algorithm3
This method combines the genetic algorithm and Otsu, and improves the performance of genetic algorithm; on the other hand, it uses the gray equalization to enhance the image before the solving process of Genetic algorithm.
- 3
- 3
- 2
- ATFU-Net: adaptive feature refinement and convolutional-style self-attention for medical image segmentation2
AdaTransFormerNet, an advanced deep learning framework addressing challenges like contrast variation, noise interference, and multi-scale anatomical structures in medical imaging, demonstrates superior performance in small-sample medical scenarios through its dual capability of global context modeling and local detail preservation.
- 2
- 2
- OncoImmune machine-learning model predicts immune response and prognosis in leiomyosarcoma2
It is found that ATRX mutation may regulate coiled-coil domain-containing protein 69 (CCDC69) expression, leading to functional alterations in mast cells and immune unresponsiveness through the modulation of various immune-related signaling pathways.
- 2
- 2
- 2
- 2
- 2
- Transferable Sequential Recommendation via Vector Quantized Meta Learning2
A vector quantized meta learning for transferable sequential recommenders (MetaRec) without requiring additional modalities or shared information across domains, which leverages user-item interactions from multiple source domains to improve the target domain performance.
- 2
- 2
- Improved Secure Lightweight RFID Authentication Protocol2
The improved protocol has meet requirements of backward security, mutual authentication, synchronization and non-denial of service, and is suitable to be deployed in mobile RFID environment and applied in low-cost RFID tags.
- Tracking by Segmentation2
Current efforts of single-target tracking mainly focus on building robust bounding box-based trackers to effectively discriminate the target from the background.
- 2
- Online Vehicle Tracking in Aerial Imagery2
This work attempts to combine different color models within the correlation-filter-based tracking framework for tracking vehicles in aerial images by exploiting a set of color models to describe the appearance of the tracked object based on correlation filters.
- 2
- 2
- Robust Online Object Tracking with a Structured Sparse Representation Model2
This study presents a novel tracking method based on the proposed structured sparse representation model, in which the tracked object is assumed to be sparsely represented by a set of object and background templates.
- TRI-TRACKING: COMBINING THREE INDEPENDENT VIEWS FOR ROBUST VISUAL TRACKING2
This paper presents a novel tri-tracking framework combining different views (different models using independent features) for robust object tracking, which exploits a hybrid discriminative generative model based on online semi-supervised learning.
- 2
- A map-matching algorithm based on analogous projection2
This paper introduces the positioning principle of mapmatching algorithm based on projection, then points out the defects of this direct projection algorithm, and then modifiest the former algorithm and puts forward a new algorithms based on analogous projection.
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- Infrared and Visible Image Fusion Algorithm Based on Threshold Segmentation1
The experimental result shows that the image fusion algorithm presented by this paper retains clear scenes of the visible image adequately, meanwhile highlights the thermal targets of the infrared image, which enriches information and object characteristics in the fusion image.
- Correction to: Deep-Sea Organisms Tracking Using Dehazing and Deep Learning1
The original version of this article unfortunately contained a mistake in the Affiliation section.
- 1
- Introduction to Visual Tracking1
This work focuses on visual tracking, a part of many high-level problems of computer vision, such as motion analysis, event detection, and activity understanding, which has been attracting increasing attention in the vision community.
- 1
- Moving space target detection algorithm based on trajectory similarity1
The motion model of the space target according to its motion characteristics is established, and a target initial screening method based on image sequence projection frame is proposed that can effectively detect space moving objects.
- Hyperspectral Image Classification via a Joint Weighted K-Nearest Neighbour Approach1
A joint KNN approach to solve the HSI classification problem by considering the distances between all neighbouring pixels of a given test pixel and training samples and exploiting a set-to-point distance between neighbouring pixels and each training sample.
- 1
- 1
- Robust Visual Tracking with Dual Group Structure1
This paper presents a robust tracking framework by exploiting the dual group structure of both candidate samples and dictionary templates, and formulate the sparse representation at group level, which is embedded into the particle filter framework for visual tracking.
- 1
- An Algorithm for Multimodal Biometrics Feature Level Fusion Recognition Based on the Second-Generation Curvelet and 2D Log-Gabor Filter1
The second generation Curvelet is employed to extract face information, and amplitudes of 2D Log-Gabor are used to extract iris information and the proposed algorithm is an effective model for multimodal biometric recognition.
- Digital Image Processing via Combination of Low-Level and High-Level Approaches.1
This thesis is organised in three parts: Colour Edge and Face Detection; Hand motion detection; Hand Gesture Detection and Medical Image Processing.
- Image Shrinkage Based on Hot-Target Map and Featured Edge Preservation1
A novel image shrinking method named as Hot-Target map to calculate quad's importance according to image saliency and edges and avoids distortion by preserving image linear edges named as featured straight edge.
- 1
- –
- Less Repetition, Less Energy Cost: A Reinforcement Learning-Based Multiagent Energy-Saving Autonomous Exploration System–
This work proposes a distributed Multiagent Energy-saving Autonomous Exploration System (MEAES), and devise the consumption-exploration-balanced training framework (CEBF), which guides agents to transform from lazy exploration to energy-saving exploration strategies through dynamic reward shaping.
- –
- –
- One Aligned LLM to Serve Them All: A Transfer Recipe for Training VLMs without Visual-Language Re-Alignment–
This work demonstrates that the aligned LLM with a general-purpose vision encoder can effectively enhance downstream VQA performance with task-specific encoders, and investigates several alignment strategies between the aligned LLM and new task-specific encoders.
- A Multi-Scale High-Resolution Gradient-Aware Convolution for Infrared Dim and Small Target Detection–
- –
- Detection of infrared moving small targets from dynamic background based on improved ViBe algorithm–
An improved ViBe algorithm incorporating adaptive threshold setting based on dynamic feedback that significantly enhances pixel classification accuracy and the real-time background model updating in diverse scenarios and improves the precision of moving target detection, reduces the false detection rates, and increases overall robustness.
- –
- –
- –
- Detection technology of moving small infrared target in dynamic background based on improved vibe algorithm–
Experimental results show that the algorithm improves Vibe’s ability to adapt to scene changes, reduces the interference of background areas on foreground areas, and can effectively detect moving small infrared targets from dynamic background.
- –
- –
- –
- A local adaptive step size template matching target tracking algorithm–
A template matching tracking algorithm with adaptive step size is proposed, which solves the problem of tracking jitter in traditional template matching tracking algorithms due to the fixed template size, which causes the target size to exceed the template at the end of the tracking.
- –
- Research on Small Target Detection Algorithm for Aerial Images based on Improved YOLOv11–
Experiments show that compared with YOLOV11n, the performance of the algorithm on small targets is greatly improved, and Wise IoU is used to replace CIoU to speed up the network convergence speed and improve the regression accuracy.
- –
- –
- Ultra-High Field 2D TOF MRA of Pelvis and Lower Extremity: Comparison of 5T and 3 T–
2D TOF MRA at 5 T is capable of providing superior vessel delineation and image quality to 3 T to demonstrate the superiority of 5T to 3T in TOF MRA for pelvis and lower extremities.
- –
- –
- –
- –
- Efficient Visual Place Recognition by Adaptive CNN Landmark Matching–
This paper proposes a simple yet effective four-step method that achieves adaptive convolutional neural network (CNN) landmark matching for VPR.
- –
- MsrFace: Multi-Sphere Radius Loss for Deep Face Recognition–
This paper proposes a Multi-Sphere Radius Loss (MsrFace) to add radius constraints and presents experiments on several widely used benchmarks to show that MsrFace has a better performance in comparison with some recent state-of-the-art face recognition methods.
- Visual Tracking Based on Sparse Representation–
Appearance change caused by intrinsic and extrinsic factors is a challenging problem in online object tracking.
- –
- Visual Tracking Based on Model Fusion–
This chapter introduces two representative methods based on model fusion, a traditional method focus on how to combine various handcraft features effectively and the second one is a deep-learning-based method which explores the attention information totally in visual tracking.
- Visual Tracking Based on Local Model–
Many effective and efficient tracking methods cannot deal with partial occlusion and background clutter, so several local-based models have been employed for designing robust tracking algorithms.
- –
- Visual Tracking Based on Deep Learning–
This chapter introduces three deep trackers (FCNT, StructSiam, ACT) and complements two convolutional layers of different levels in handling drastic appearance change and distinguishing target object from its similar distractors.
- Image Fusion Method of Infrared and Optical Images Based on Target Extraction–
Experimental results indicate that the proposed fusion method outperforms traditional fusion methods in preserving infrared target information in the fused image.
- –
- Online object tracking via bag-of-local-patches–
This paper develops a novel tracking algorithm based on the bag- of-local-patches representation with the discriminative learning scheme that performs better than other state-of-the-art tracking methods.
- –
- –
- –
- –
- GUI Video and Image Processing of UAV Reconnaissance Videos Based on Matlab–
Simulation results indicate that the proposed GUI is promising in the field of video and image processing concerning UAV reconnaissance videos.
- –
- Research on the Detection and Identification Technology of Early Oil Fire Images–
An algorithm of early oil fire image detection and recognition is put forward and the results of the simulated tunnel experiments show that the algorithm can successfully detect and recognize oil fire.
- –
- A Colorization Algorithm for Gray Images Based on Image Segmentation and Color Propagation–
This paper introduces a technique for colorizing the grayscale images by segmenting the source image and the target image using the Mean-Shift algorithm to produce a fully colorized target image.
- Analysis of Transportation Need Estimate on Land Uses–
The article introduced the model of the transportation need estimate and categorize to provide a kind of way of combining land uses program and transportation program.
- Port-Based RADIUS Authentication Management and Accounting System Design–
This article first introduces two traditional user access authentication technology (PPPoE, the WEB authentication) limitation, then detailedly introduces 802.1X port-based network access control technology, andaborates the IEEE802.1x authentication management principle and the authentication process.
- –
- –
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-14. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.