Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks141 from public data
- CenterNet: Keypoint Triplets for Object Detection3,587
This paper presents an efficient solution that explores the visual patterns within individual cropped regions with minimal costs, and builds the framework upon a representative one-stage keypoint-based detector named CornerNet, which improves both precision and recall.
- CenterNet: Object Detection with Keypoint Triplets3,563
This paper presents an efficient solution that explores the visual patterns within individual cropped regions with minimal costs, and builds the framework upon a representative one-stage keypoint-based detector named CornerNet, which improves both precision and recall.
- Celeb-DF: A Large-Scale Challenging Dataset for DeepFake Forensics1,885
This work presents a new large-scale challenging DeepFake video dataset, Celeb-DF, which contains 5,639 high-quality DeepFake videos of celebrities generated using improved synthesis process and conducts a comprehensive evaluation of DeepFake detection methods and datasets to demonstrate the escalated level of challenges posed by Celebrity-DF.
- UA-DETRAC: A new benchmark and protocol for multi-object detection and tracking799
This work performs a comprehensive quantitative study on the effects of object detection accuracy to the overall MOT performance, using the new large-scale University at Albany DETection and tRACking (UA-DETRAC) benchmark dataset.
- The Visual Object Tracking VOT2016 Challenge Results776
The Visual Object Tracking challenge VOT2016 goes beyond its predecessors by introducing a new semi-automatic ground truth bounding box annotation methodology and extending the evaluation system with the no-reset experiment.
- Multi-Scale Structure-Aware Network for Human Pose Estimation300
A robust multi-scale structure-aware neural network for human pose estimation that effectively improves state-of-the-art pose estimation methods that suffer from difficulties in scale varieties, occlusions, and complex multi-person scenarios.
- A dual‐stage attention‐based Conv‐LSTM network for spatio‐temporal correlation and multivariate time series prediction241
A convolution LSTM network model based on MTS prediction with two‐stage attention that is not only suitable for single‐step prediction of MTS, but also suitable for multistep prediction of time step in a certain range.
- Celeb-DF: A New Dataset for DeepFake Forensics214
A new DeepFake dataset is presented, Celeb-DF, for the development and evaluation of DeepFake detection algorithms, generated using a refined synthesis algorithm that reduces the visual artifacts observed in existing datasets.
- 145
- One-Class Classification with Extreme Learning Machine145
The experimental evaluation shows that the ELM based one-class classifier can learn hundreds of times faster than autoencoder and it is competitive over a variety of one- class classification methods.
- VisDrone-DET2018: The Vision Meets Drone Object Detection in Image Challenge Results133
A large-scale drone-based dataset, including 8, 599 images with rich annotations, including object bounding boxes, object categories, occlusion, truncation ratios, etc, is released, to narrow the gap between current object detection performance and the real-world requirements.
- Exposing GAN-synthesized Faces Using Landmark Locations127
This work describes a new method to expose GAN-synthesized images using the locations of the facial landmark points, based on the observations that the facial parts configuration generated by GAN models are different from those of the real faces, due to the lack of global constraints.
- Corner Proposal Network for Anchor-Free, Two-Stage Object Detection121
A novel anchor-free, two-stage framework which first extracts a number of object proposals by finding potential corner keypoint combinations and then assigns a class label to each proposal by a standalone classification stage is proposed, which is competitive among state-of-the-art object detection methods.
- CenterNet++ for Object Detection112
It is demonstrated that bottom-up approaches show competitive performance compared with top-down approaches and have higher recall rates and a real-time CenterNet model, an anchor-free detector that achieves a good trade-off between accuracy and speed.
- The importance of short lag-time in the runoff forecasting model based on long short-term memory98
The experimental results demonstrate that, compared with state-of-art benchmark models, SA-LSTM achieves the best performance and can be attributed to its exploitation of information in short lag-time.
- A novel loss function of deep learning in wind speed forecasting97
A kernel MSE loss function to evaluate the ubiquitous nonlinearity of deep learning errors in the reproducing kernel Hilbert space is proposed and the results imply that developing a loss function with kernel skills is a new way to get better results.
- 93
- 90
- Skill-SD: Skill-Conditioned Self-Distillation for Multi-turn LLM Agents80
Skill-SD is introduced, a framework that turns the agent's own trajectories into dynamic training-only supervision and substantially outperforms the standard RL baseline, improving both vanilla GRPO and vanilla OPD on agentic benchmarks.
- 66
- The Thermal Infrared Visual Object Tracking VOT-TIR2016 Challenge Results66
A significant general improvement of results has been observed, which partly compensate for the more difficult sequences introduced into the dataset in VOT-TIR2016.
- Multi-Camera Multi-Target Tracking with Space-Time-View Hyper-graph66
A new multi-camera multi-target tracking method based on a space-time-view hyper-graph that encodes higher-order constraints on 3D geometry, appearance, motion continuity, and trajectory smoothness among 2D tracklets within and across different camera views is proposed.
- DETRAC: A New Benchmark and Protocol for Multi-Object Tracking63
This work performs a comprehensive quantitative study on the effect of object detection accuracy to the overall MOT performance based on a new large-scale DETection and tRACking (DETRAC) benchmark dataset.
- Geometric Hypergraph Learning for Visual Tracking58
A geometric hypergraph learning-based tracking method, which fully exploits high-order geometric relations among multiple correspondences of parts in different frames, and a confidence-aware sampling method is developed to select representative vertices and hyperedges to construct the geometric hyper graph for more robustness and scalability.
- 55
- 50
- Multi-Scale Supervised Network for Human Pose Estimation45
A robust pose estimation method based on the recent deep conv-deconv modules with two improvements: (1) multi -scale supervision of body keypoints, and (2) a global regression to improve structural consistency of keypoints.
- 45
- Contrast Enhancement Estimation for Digital Image Forensics41
The nature of contrast enhancement as a mapping between pixel values and the distinct characteristics it introduces to the image pixel histogram is taken advantage of, and a new method to estimate contrast enhancement operations from a single image is described.
- MVDR Algorithm Based on Estimated Diagonal Loading for Beamforming39
This work improves MVDR algorithm based on estimated diagonal loading for beamforming by establishing a model based on diagonal loading compensation and determining the optimal diagonal loading value in the interval through the experimental method.
- Pulmonary nodules detection assistant platform: An effective computer aided system for early pulmonary nodules detection in physical examination33
Experimental results show that the proposed pulmonary nodule detection model is robust for different datasets, which can successfully detect smaller and larger nodules in CT images obtained by physical examination.
- 33
- 31
- 28
- LandmarkGAN: Synthesizing faces from landmarks25
This work describes a new method, namely LandmarkGAN, to synthesize faces based on facial landmarks as input, able to transform a set of facial landmarks into new faces of different subjects, while retains the same facial expression and orientation.
- 25
- 25
- Landmark Breaker: Obstructing DeepFake By Disturbing Landmark Extraction25
Landmark Breaker is described, the first dedicated method to disrupt facial landmark extraction, and applied to the obstruction of the generation of DeepFake videos, achieved using adversarial perturbations.
- 25
- 25
- DeepFake-o-meter: An Open Platform for DeepFake Detection24
This work develops an open-source online platform, known as DeepFake-o-meter, that integrates state-of-the-art DeepFake detection methods and provide a convenient interface for the users.
- 24
- 23
- 20
- 19
- An Intelligent prediction model for UCG state based on dual-source LSTM19
The paper proposes the concept of combustible gas equivalents that can characterize the concentration of mixed gas through stoichiometry and material balance equations and shows that compared with Support Vector Machine (SVM) and Back Propagation Neural Network (BPNN) prediction model, the model can make a better prediction of equivalent value.
- The optimal game model of energy consumption for nodes cooperation in WSN19
The experimental results show that the optimal game model of energy consumption for nodes cooperation has obvious advantages in energy-saving efficiency and the Nash Equilibrium of cooperative game is solved using the nodes income distribution method based on Shapley value.
- COMICS: End-to-End Bi-Grained Contrastive Learning for Multi-Face Forgery Detection18
This paper proposes Contrastive Multi-FaceForensics (COMICS), an end-to-end framework for multi-face forgery detection that integrates face extraction and forgery detection in a seamless manner and adapts to the advanced object detection architectures.
- 18
- FREEWAY INCIDENT DETECTION USING KINEMATIC DATA FROM PROBE VEHICLES17
An incident detection algorithm based on the speed and acceleration profiles of probe vehicles as they travel along a freeway that can achieve comparable detection rate and mean time to detection against the neural network model is presented.
- 16
- 14
- Facial Landmarks Detection by Self-Iterative Regression Based Landmarks-Attention Network13
This paper develops a Self-Iterative Regression (SIR) framework, which concurrently learns features around each landmark and obtains the holistic location increment, and proposed Landmarks-Attention Network (LAN) as the authors' regressor.
- A Dynamic Travel Time Estimation Model Based on Connected Vehicles13
A dynamic travel time estimation model is presented which can collect and distribute traffic data based on the connected vehicles and a road link dynamic dividing algorithm is proposed to estimate the real-time travel time more accurately.
- A spatiotemporal weighted dissimilarity-based method for video saliency detection13
Developing models that lead to motion features that are extracted from videos and used in a new video saliency detection method called spatial-temporal weighted dissimilarity (STWD), which is highly competitive with, and delivers better performance than state-of-the-art methods.
- A Bio-Inspired QoS-Oriented Handover Model in Heterogeneous Wireless Networks13
The comparative analysis shows that the bio-inspired model outperforms the utility function based handover decision scheme in terms of ensuring a better QoS satisfaction and a better fairness of network resources allocation in dynamic heterogeneous wireless networks.
- Generalized Framework for Similarity Measure of Time Series13
This paper proposes a generalized framework to measure the similarity of time series, and it can be claimed that the similarity measure functions of linear multivariable time series based on the norm distance of covariance matrix and nonlinear multivariables time seriesbased on kernel function are reasonable and practical.
- DGYOLOv8: An Enhanced Model for Steel Surface Defect Detection Based on YOLOv812
An enhanced YOLOv8 model (DCNv4_C2f + GAM + GAM + InnerMPDIoU + YOLOv8, hereafter referred to as DGYOLOv8) is developed to tackle the challenges of object detection in steel surface defect detection tasks.
- FakeTracer: Catching Face-Swap DeepFakes via Implanting Traces in Training11
A new proactive defense method called FakeTracer is described to expose face-swap DeepFakes via implanting traces in training via implanting traces in training and extensive experiments corroborate the efficacy of this method on defending against face-swap DeepFake.
- 11
- Toward the Creation and Obstruction of DeepFakes11
A new large-scale challenging DeepFake video dataset, Celeb-DF, is presented, which contains 5, 639 high-quality DeepFake videos of celebrities generated using an improved synthesis process, and Landmark Breaker, the first dedicated method to disrupt facial landmark extraction, is introduced.
- 11
- 10
- 9
- 9
- 9
- 9
- 9
- 9
- Reconfigurable video coding framework and decoder reconfiguration instantiation of AVS9
This paper presents the procedure of describing a reconfigured decoder in DD, reusing and exchanging tools from VTLs and initializing ADM in the dataflow formalism of RVC framework, and shows that the process mechanism offered by R VC framework is versatile and flexible to achieve high reusability and exchangeability in decoder configurations.
- Hints of Prompt: Enhancing Visual Representation for Multimodal LLMs in Autonomous Driving8
The Hints of Prompt framework is proposed, which introduces three key enhancements: Affinity hint to emphasize instance-level structure by strengthening token-wise connections, Semantic hint to incorporate high-level information relevant to driving-specific cases, and Question hint to align visual features with the query context, focusing on question-relevant regions.
- 8
- 7
- TransRPN: Towards the Transferable Adversarial Perturbations using Region Proposal Networks and Beyond7
As RPNs have no constraint on size of input image, the TransRPN method can generate the adversarial images directly into object detectors with arbitrary input size, which improves the feasibility of the method in practical applications.
- Improving Image Restoration with Soft-Rounding7
This work introduces a new regularizer that attains zero at the designated pixel values and becomes a quadratic penalty function in the intervals between them and applies this regularizer to the restoration of binary text/barcode images and pattern images with multiple distinct pixel values.
- Fast Online Upper Body Pose Estimation from Video7
This work describes a fast online video upper body pose estimation method (CDBN-MODEC) that is based on a conditional dynamic Bayesian network model, which predicts upper body poses in a frame without using information from future frames.
- An Efficient Multitask Scheduling Model for Wireless Sensor Networks7
By applying the proposed priority based data processing model in the multitask wireless sensor network, the average delay of data packets in a sensor nodes is reduced nearly to 50% and the simulation results show that the proposed model can improve the throughput of network efficiently.
- 7
- 7
- 6
- Dense-scale dynamic network with filter-varying atrous convolution for semantic segmentation6
A filter-varying atrous convolution (FAC) to efficiently enlarge the per-pixel receptive fields pertaining to various objects and a dense-scale dynamic network (DsDNet) that simultaneously enhances the localization accuracy and reduces the effect of large-scale variations of the object, by assigning FAC to different spatial locations at dense scales.
- 3D Video Rendering Adaptation: A Survey6
This paper surveys the state-of-the-art 3D video rendering adaption technologies and further summarizes the remaining open challenges and explains the basic 3D vision principle which provides 3DVideo rendering adaptation rules.
- 6
- 6
- Ground-Glass Lung Nodules Recognition Based on CatBoost Feature Selection and Stacking Ensemble Learning5
Experiments show that the proposed recognition method based on CatBoost feature selection and Stacking ensemble learning can effectively select features and make recognition on ground-glass pulmonary nodules.
- 5
- An FPGA-friendly CABAC-encoding architecture with dataflow modelling programming5
An optimized ‘Producer–Consumer’ architecture for CABAC by dataflow modelling is proposed and the buffer accessing speed and buffer size in the architecture is analysed and refined to achieve high-throughput and low-resource consumption.
- 5
- 5
- 4
- Algorithm analysis and architecture design for rate distortion optimized mode decision in high definition AVS video encoder4
Hardware oriented mode decision algorithm is tailored for VLSI implementation in this work for high definition video coding, and intelligent pipeline scheduling mechanism is proposed to break the intrinsic data dependency in intra prediction, which is directly related with mode decision.
- Utilization of CNN-LSTM Model in Prediction of Multivariate Time Series for UCG3
The CNN-LSTM model is proposed, which reconstructs the UCG concentration data so that they represent spatial and temporal properties, respectively, and outperforms prior methods.
- Semi-supervised class-specific feature selection for VHR remote sensing images3
A weight matrix fully integrates local geometrical structure and discriminative information and is incorporated into a -norm minimization optimization problem of data reconstruction to objectively measure the effectiveness of features for a thematic class.
- A Novel Smart Multi-license Plate Recognition Method3
This paper presents a Smart License Plate Recognition System (SLPRS) which employs a cascade scheme guaranteeing high accuracy and reliability and adopts pattern match method for recognition based on proper designed character templates.
- Visual Saliency and Distortion Weighting Based Video Quality Assessment3
In this paper, a new model called saliency and distortion weighted structural similarity index with temporal pooling strategy (SDTW-SSIM) is proposed, which performs well on both computational efficiency and assessment accuracy.
- 3
- Error-resistance and Low-complexity Integer Inverse Discrete Cosine Transform3
A low-complexity multiplication-free integer inverse discrete cosine transform based on data flow structure of improved AAN’s fast IDCT is proposed for error drifting resistance of decoder and shows that it reduces the error drifting of decoders efficiently.
- 3
- MEMTS: Internalizing Domain Knowledge via Parameterized Memory for Retrieval-Free Domain Adaptation of Time Series Foundation Models2
MEMTS is proposed, a lightweight and plug-and-play method for retrieval-free domain adaptation in time series forecasting that enables MEMTS to achieve accurate domain adaptation with constant-time inference and near-zero latency, while effectively mitigating catastrophic forgetting of general temporal patterns.
- 2
- 2
- 2
- A Regular VLSI Architecture of Motion Vector Prediction for Multiple-Standard MPEG-Like Video Codec2
A highly regular architecture to implement MV prediction for multi-standard video codec is proposed and pipelining and parallelism are employed in the proposed architecture to improve throughput efficiency and tradeoff between hardware cost and efficiency.
- 2
- HER2 Expression Prediction with Flexible Multi-Modal Inputs via Dynamic Bidirectional Reconstruction1
An adaptive bimodal input prediction framework that flexibly supports both single-modality and dual-modality inputs, enabling accurate predictions using either H&E or IHC images alone, while retaining the ability for joint inference when both modalities are available.
- A Diffusion Model Based Quality Enhancement Method for HEVC Compressed Video1
Compared with the existing methods, this method requires only one model to improve the quality of videos encoded with different quantization parameters, and the structure of this model is shown in Figure 1.
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- A study on the motion vector prediction schemes for AVS1
Two MVP schemes for a new coding standard, Audio and Video Standard (AVS), are proposed, suitable for different video sequences with different motion characteristics, respectively, and the experimental results show that the bit rate savings are achieved with these schemes, compared with the median predictor implemented in AVS.
- –
- –
- A generalizable eye disease detection method based on Zero-Shot Learning–
ZSL can simulate clinical diagnostic logic and generalize to unseen eye diseases, offering a promising approach for automated screening where labeled data are scarce and outperforming most supervised baselines on external test datasets.
- –
- BRIDGE: Behavior-Guided Residual Integration with Dual-Frequency Graph Evidence–
BRIDGE, a behavior-guided residual integration framework built on dual-frequency graph evidence, is proposed, which keeps the multimodal backbone and localizes behavior evidence to candidate calibration.
- GraSP-VL: Length as a Semantic Granularity Interface for Vision-Language Representations–
The results show that frozen VLM embeddings can be reorganized into a truncatable semantic prefix interface rather than merely compressed.
- –
- RetinaRegen: A Hybrid Model for Readability and Detail Restoration in Fundus Images–
RetinaRegen is proposed, a hybrid model for retinal image restoration that integrates a readability classifi-cation model, a Diffusion Model, and a Variational Autoencoder (VAE), offering an effective solution to enhance fundus image quality and support clinical diagnosis.
- –
- RURA-Net: A general disease diagnosis method based on Zero-Shot Learning–
This study proposes a general disease diagnosis approach based on Zero-Shot Learning, and the Siamese neural network is used to find similar diseases for the target diseases, and the U-Net segmentation model is used to accurately segment the key lesions of the disease.
- Three-Stage Multi-Frame Multi-Channel In-Loop Filter of VVC–
A novel three-stage Multi-frame Multi-channel In-loop Filtering (3-MMIF) method for VVC that improves the quality of each encoded frame by harnessing the correlations between adjacent frames and channels and leveraging spatio-temporal and inter-channel correlations.
- DynSegNet:Dynamic Architecture Adjustment for Adversarial Learning in Segmenting Hemorrhagic Lesions from Fundus Images–
This paper proposes an adversarial learning-based dynamic architecture adjustment approach that integrates hierarchical U-shaped encoder-decoder, residual blocks, attention mechanisms, and ASPP modules that enhances segmentation performance by dynamically optimizing feature fusion.
- –
- Active Fake: DeepFake Camouflage–
This work introduces a new framework for creating DeepFake camouflage that generates blending inconsistencies while ensuring imperceptibility, effectiveness, and transferability, and crafts imperceptible yet effective inconsistencies to mislead forensic detectors.
- Automatic Block-wise Pruning with Auxiliary Gating Structures for Deep Convolutional Neural Networks–
A novel structured network pruning method with auxiliary gating structures which assigns importance marks to blocks in back-bone network as a criterion when pruning, thus achieving a better performance than the unpruned model with over 93% FLOPs re-duced.
- –
- –
- –
- Spatio-temporal Visual Distortion and Rate Optimization for Video Coding–
A rate-distortion model based on spatio-temporal video structural similarity (stVSSIM) index, which takes both spatial and temporal visual quality into account is proposed, which averagely reduces 20% bit-rate under the equal visual quality and the adaptive Lagrange multiplier can further improve the results.
- High throughput VLSI architecture for multiresolution integer motion estimation in high definition AVS video encoder–
By parallel searching and utilizing the high correlation in multi-resolution reference pixels, huge throughput and computation due to large search window are alleviated considerably and 16 way parallel processing element arrays with configurable multiplying technologies achieve fast search with regular data access and efficient data reuse.
- Implementation of Large Number of Buildings Based–
A highly efficient large-scale high-density complex 3D graphics engine based on OpenGL based on Visual C++ 2005 and OpenGL is built, which realizes the three dimensional real world lifelike return to original state.
- Analysis of AVS Video Conformance Testing–
The definition and general methods for the conformance testing of AVS-coded video bitstream are summarized and the design of some conformance bitstreams are discussed.
- –
- PRE-Processing for Video Coduing with Rate-Distortion Optimization Decision–
This paper proposes an adaptive video pre-processing algorithm that adopts Gaussian filter to remove noise and insignificant features existing in images of video and can be compliant with any video coding standards without modifying the decoder.
- –
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.