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
- 496
- Wing Loss for Robust Facial Landmark Localisation with Convolutional Neural Networks479
A new loss function, namely Wing loss, for robust facial landmark localisation with Convolutional Neural Networks (CNNs) is presented, and the superiority of the proposed method over the state-of-the-art approaches is proved.
- 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.
- Joint Group Feature Selection and Discriminative Filter Learning for Robust Visual Object Tracking222
It is demonstrated that the GFS-DCF method is able to significantly improve the performance of a DCF tracker equipped with deep neural network features and enables joint feature selection and filter learning, achieving enhanced discrimination and interpretability of the learned filters.
- MECT: Multi-Metadata Embedding based Cross-Transformer for Chinese Named Entity Recognition143
A novel Multi-metadata Embedding based Cross-Transformer (MECT) to improve the performance of Chinese NER by fusing the structural information of Chinese characters by using multi-metadata embedding in a two-stream Transformer.
- Vision-Based UAV Self-Positioning in Low-Altitude Urban Environments140
A new dataset, DenseUAV, that is the first publicly available dataset tailored for the UAV self-positioning task and incorporates metric learning into representation learning to enhance the model’s discriminative capacity and to reduce the modality discrepancy.
- 130
- 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.
- Dynamic Attention-Controlled Cascaded Shape Regression Exploiting Training Data Augmentation and Fuzzy-Set Sample Weighting105
A new Cascaded Shape Regression architecture, namely Dynamic Attention-Controlled CSR (DAC-CSR), for robust facial landmark detection on unconstrained faces with fault-tolerant mechanism using fuzzy set sample weighting for attention-controlled domain-specific model training.
- 99
- One Model for ALL: Low-Level Task Interaction Is a Key to Task-Agnostic Image Fusion91
This new paradigm provides strong guidance for unsupervised multimodal fusion without relying on abstract semantics, enhancing task-shared feature learning for broader applicability.
- MFR-DTA: a multi-functional and robust model for predicting drug–target binding affinity and region89
A novel Multi-Functional and Robust Drug–Target binding Affinity prediction (MFR-DTA) method that achieves excellent performance on these datasets, demonstrating its merits and superiority over the state-of-the-art methods.
- 86
- Gaussian mixture 3D morphable face model85
Experiments in fitting the GM-3DMM to 2D face images to facilitate their geometric and photometric normalisation for pose and illumination invariant face recognition demonstrate the merits of the proposed mixture of Gaussians 3D face model.
- MMDG-DTI: Drug–target interaction prediction via multimodal feature fusion and domain generalization80
- 76
- Adaptive Channel Selection for Robust Visual Object Tracking with Discriminative Correlation Filters72
The robustness and stability of the learned appearance model are significantly enhanced by the proposed method as the process of channel selection performs implicit spatial regularisation, using the augmented Lagrangian method to optimise the discriminative filters efficiently.
- 67
- Evaluation of Dense 3D Reconstruction from 2D Face Images in the Wild67
This paper investigates the evaluation of dense 3D face reconstruction from a single 2D image in the wild by organising a competition that provides a new benchmark dataset that contains 2000 2D facial images of 135 subjects as well as their 3D ground truth face scans.
- 58
- 58
- 53
- 50
- 45
- Face Detection, Bounding Box Aggregation and Pose Estimation for Robust Facial Landmark Localisation in the Wild44
The experimental results obtained on the 300W and Menpo benchmarks demonstrate the superiority of the robust face detection and landmark localisation framework over state-of-the-art methods.
- 43
- 43
- Efficient 3D morphable face model fitting42
An efficient stepwise 3DMM-to-2D image-fitting procedure, which sequentially optimises the pose, shape, light direction, light strength and skin texture parameters in separate steps leading to an accurate and efficient fitting is presented.
- Rectified Wing Loss for Efficient and Robust Facial Landmark Localisation with Convolutional Neural Networks40
This paper presents a new loss function, namely Rectified Wing (RWing) loss, for regression-based facial landmark localisation with Convolutional Neural Networks (CNNs), and proposes a simple but effective boosting strategy, referred to as pose-based data balancing, for under-representation of samples with large pose variations.
- NFLAT: Non-Flat-Lattice Transformer for Chinese Named Entity Recognition38
A novel lexical enhancement method, InterFormer, that effectively reduces the amount of computational and memory costs by constructing non-flat lattices is advocated and implemented, with InterFormer as the backbone, for Chinese NER.
- Dynamic dictionary optimization for sparse-representation-based face classification using local difference images36
This study presents a new sparse-representation-based face-classification algorithm that exploits dynamic dictionary optimization on an extended dictionary using synthesized faces and uses a dictionary-optimization method to reduce the redundancy and improve the classification accuracy.
- 3D Morphable Face Models and Their Applications36
Recent developments in 3D face modelling and model fitting will be overviewed, and their merits in the context of diverse applications illustrated on several examples, including pose and illumination invariant face recognition, and3D face reconstruction from video.
- 34
- An accelerated correlation filter tracker34
This work proposes an accelerated ADMM optimisation method obtained by adding a momentum to the optimisation sequence iterates, and by relaxing the impact of the error between DCF parameters and their norm, which achieves significantly increased convergence of the DCF filter.
- 33
- 32
- 32
- 31
- Collaborative representation based face classification exploiting block weighted LBP and analysis dictionary learning31
A block weighted LBP histogram algorithm is designed to form a set of local histogram-based feature vectors instead of using raw images to effectively decrease data redundancy and uncertainty derived from image noises and appearance variations.
- Dictionary Integration Using 3D Morphable Face Models for Pose-Invariant Collaborative-Representation-Based Classification29
A dictionary integration algorithm using 3D morphable face models (3DMM) for pose-invariant collaborative-representation-based face classification and an on-line pruning strategy under the CRC framework is presented.
- Reinforcement Learning for Scalable Train Timetable Rescheduling With Graph Representation28
A reinforcement learning-based approach to TTR, which design a simple directed graph to represent the TTR problem, enabling the automatic extraction of informative states through graph neural networks, and designs a learning curriculum for the model to handle the scenarios with different levels of delay.
- 26
- 26
- 26
- 24
- Dual Encoder-Decoder Based Generative Adversarial Networks for Disentangled Facial Representation Learning21
An extensive quantitative and qualitative evaluation carried out on several controlled and in- the-wild benchmarking datasets demonstrates the superiority of the proposed DED-GAN method over the state-of-the-art approaches.
- Half-Face Dictionary Integration for Representation-Based Classification21
The proposed HFDI algorithm measures residuals between an input signal and the reconstructed one, using both the original and the synthesized dual-column (row) half-face training samples, and proposes an elimination strategy to gain the most robust training atoms.
- Revisiting RGBT Tracking Benchmarks From the Perspective of Modality Validity: A New Benchmark, Problem, and Solution18
This work presents a new benchmark considering the modality validity, MV-RGBT, captured specifically from MMW scenarios where either RGB or TIR modality is invalid, and proposes a new solution with a mixture of experts, named MoETrack, where each expert generates independent tracking results along with a confidence score.
- Efficient Few-Shot Action Recognition via Multi-level Post-reasoning18
By design, a post-reasoning mechanism is proposed for domain adaptation, which avoids most gradient backpropagation, improving the efficiency; meanwhile, a multi-level representation is utilised during the reasoning and matching processes to improve the discriminability, ensuring effectiveness.
- 18
- 18
- 17
- 17
- 15
- PropVG: End-To-End Proposal-Driven Visual Grounding with Multi-Granularity Discrimination14
This work proposes PropVG, an end-to-end proposal-based framework that is the first to seamlessly integrate foreground object proposal generation with referential object comprehension without requiring additional detectors, and introduces a Contrastive-based Refer Scoring module.
- 14
- A Flatter Loss for Bias Mitigation in Cross-dataset Facial Age Estimation13
This paper proposes a novel loss function that is more effective for neural network training that leads to the convergence of SGD to a better optimum point, and consequently a better generalisation in facial age estimation.
- Improving Generalized Visual Grounding With Instance-Aware Joint Learning12
InstanceVG is the first framework to simultaneously tackle both GREC and GRES while incorporating instance-aware capabilities into generalized visual grounding, and is the first framework to simultaneously tackle both GREC and GRES while incorporating instance-aware capabilities into generalized visual grounding.
- 10
- Learning Feature Restoration Transformer for Robust Dehazing Visual Object Tracking10
This work designs a feature restoration transformer that adaptively enhances the representation capability of the extracted visual features for robust tracking in both natural and hazy scenarios, and endow the feature extraction process with a refinement mechanism typically for hazy imaging scenarios, establishing a tracking system that is robust against foggy videos.
- 10
- 10
- 9
- 9
- A Unified Tensor-based Active Appearance Model9
This work presents a Unified Tensor-based Active Appearance Model (UT-AAM) for jointly modelling the geometry and texture information of 2D faces and demonstrates the improvements in training facial landmark detectors through the use of the model to synthesise a large number of virtual samples.
- R-DTI: Drug Target Interaction Prediction Based on Second-Order Relevance Exploration8
A novel R-DTI method by exploring the second-order relevance in both protein structural feature extraction and DTI prediction phases is proposed, and an inter-feature structure-preserved Riemannian network is designed to expand the existing protein extraction patterns.
- Rethinking Positive Pairs in Contrastive Learning8
This paper presents SimLAP: a simple framework for learning visual representation from arbitrary pairs, which optimises the similarity of an arbitrary pair of samples, while simultaneously learning the enabling subspace.
- 8
- View-shuffled clustering via the modified Hungarian algorithm8
This work proposes a novel multi-view clustering method, namely View-shuffled Clustering via the Modified Hungarian Algorithm (VsC-mH), which establishes the cross-view correspondence of the shuffled data utilizing strategies of the global alignment and modified Hungarian algorithm based intra-category alignment and generates the partition of the aligned data employing matrix factorization.
- 8
- LabelPrompt: Effective Prompt-based Learning for Relation Classification8
This paper presents a novel prompt-based learning method, namely LabelPrompt, for the relation classification task, motivated by the intuition to ``GIVE MODEL CHOICES!'', which defines additional tokens to represent relation labels and implements an entity-aware module with contrastive learning.
- 8
- DGFM: Full Body Dance Generation Driven by Music Foundation Models7
This work proposes a diffusion-based method that generates dance movements conditioned on text and music that extracts music features by combining high-level features obtained by music foundation model with hand-crafted features, thereby enhancing the quality of generated dance sequences.
- PortraitTalk: Towards Customizable One-Shot Audio-to-Talking Face Generation7
This work introduces a novel, customizable one-shot audio-driven talking face generation framework, named PortraitTalk, which utilizes a latent diffusion framework consisting of two main components: IdentityNet and AnimateNet, designed to preserve identity features consistently across the generated video frames.
- Soul: An OCTA dataset based on Human Machine Collaborative Annotation Framework7
A BRVO-focused dataset named Soul (Source of ocular vascular) is developed and a human machine collaborative annotation framework (HMCAF) using scrambled retinal blood vessels data is proposed using scrambled retinal blood vessels data.
- Attention-based investigation and solution to the trade-off issue of adversarial training7
An Attention-Enhanced Learning Framework for robustness training is introduced to enable the model to inherit the attention pattern of standard pre-trained model through an embedding-level regularization, and shows remarkable improvements in generalization under the high-level robustness.
- 7
- Global Context-Aware Feature Extraction and Visible Feature Enhancement for Occlusion-Invariant Pedestrian Detection in Crowded Scenes7
A Global Context-Aware module to exploit contextual information for pedestrian detection by fusing global context with the information derived from the visible part of occluded pedestrians enhances feature representations.
- 7
- 7
- Biased Feature Learning for Occlusion Invariant Face Recognition7
A Biased Feature Learning framework for occlusion-invariant face recognition that enhances the robustness of a network to unknown occlusions but also maintains or even improves its performance for normal faces.
- Advances in facial landmark detection7
The challenges and latest technology advances in facial landmarks are looked at, including face recognition, emotion analysis, human-computer interaction and 3D face reconstruction.
- 6
- Towards multi-scale fuzzy sparse discriminant analysis using local third-order tensor model of face images6
A multi-scale fuzzy sparse discriminant analysis using a local third-order tensor model of face images to perform robust face classification and evaluates the sparse similarity grade between each training sample and class, and redesign the traditional discriminant criterion that contains considerable fuzzy sparse similarity grades.
- An Improved AAM Method for Extracting Human Facial Features6
Experimental results show that the improved algorithm can increase the accuracy of the AAM fitting and express more information for structures of edge and texture.
- 5
- Fast SRC using quadratic optimisation in downsized coefficient solution subspace5
Experimental results conducted on a set of well-known face datasets, including AR, FERET, LFW and FRGC, demonstrate the merits of the proposed methods, especially in computational efficiency.
- Feature Level Multiple Model Fusion Using Multilinear Subspace Analysis with Incomplete Training Set and Its Application to Face Image Analysis5
This paper develops a common theoretical framework for multiple model fusion at the feature level using multilinear subspace analysis (also known as tensor algebra) and applies the proposed framework to the problem of face image analysis using Active Appearance Model to validate its performance.
- 4
- 4
- Beyond Accuracy: Statistical Measures and Benchmark for Evaluation of Representation from Self-Supervised Learning4
A large-scale benchmark with diversity and granularity of classes, Statistical Metric Learning Benchmark (SMLB) built upon ImageNet-21K and WordNet is introduced, revealing the limitations of supervised learning and the class bias inherent in SSL models and offering insights into potential areas for future model enhancement.
- ETM-face: effective training sample selection and multi-scale feature learning for face detection4
An Improved Training Sample Selection (ITSS) strategy for mining effective positive and negative samples during network training and a Residual Feature Pyramid Fusion module that collects semantically robust features to improve the scale-invariance of deep features and better represent faces at different feature pyramid levels are proposed.
- Face anti-spoofing with local difference network and binary facial mask supervision4
A two-stream network that consists of a convolutional network and a local difference network, aiming for discriminative information mining and information redundancy elimination for face anti-spoofing systems and a simple binary facial mask supervision strategy is advocated.
- 4
- Learning a representation with the block-diagonal structure for pattern classification4
An approach that learns representation with block-diagonal structure (RBDS) for robust image recognition and the experimental results obtained on several benchmarking datasets demonstrate the efficacy of the proposed RBDS method.
- Improve the Spoofing Resistance of Multimodal Verification with Representation-Based Measures4
The experimental results on a chimeric multimodal database with face and ear traits demonstrate that the proposed representation-based measure significantly outperforms the well-known fusion methods with conventional measure.
- 4
- Progressive sparse representation-based classification using local discrete cosine transform evaluation for image recognition4
The final goal of the proposed method is to determine an optimal weighted sum of nearest neighbors that are obtained under the local correlative degree evaluation, which is approximately equal to the test sample, and it can use this weighted linear combination to perform robust classification.
- 3
- 3
- 3
- Investigating Self-Supervised Methods for Label-Efficient Learning3
This work conducts a systematic examination of different self-supervised pretext tasks, namely contrastive learning, clustering, and masked image modelling, to assess their low-shot capabilities by comparing different pretrained models and introduces a framework that combines mask image modelling and clustering as pretext tasks.
- 3
- 3
- 3
- Pseudo Labelling for Enhanced Masked Autoencoders3
This paper proposes an enhanced approach that boosts MAE performance by integrating pseudo labelling for both class and data tokens, alongside replacing the traditional pixel-level reconstruction with token-level reconstruction.
- 3
- 3
- Research on Reliability Evaluation of Scissor Lifting Structure Based on BP Neural Network3
A method based on neural network for reliability evaluation of scissor lift mechanism is verified, which proves the feasibility of the neural network applied to the reliability evaluation, and provides a new method and idea for the reliability Evaluation of the scissur lift mechanism.
- GCDance: Genre-Controlled Music-Driven 3D Full Body Dance Generation2
A text-based control mechanism is developed that maps input prompts into genre-specific control signals, enabling precise and controllable text-guided generation of genre-consistent dance motions, and the superiority of GCDance over the existing state-of-the-art approaches is demonstrated.
- 2
- 2
- Robust facial landmark detection in the wild2
This thesis studies robust facial landmark detection algorithms for faces in the wild and proposes the use of tensor completion methods to reconstruct missing shape and global texture in tensor-based active appearance model and develops an adaptive scheme for scale-invariant FLD.
- 2
- DASViT: Differentiable Architecture Search for Vision Transformer1
Differentiable Architecture Search for Vision Transformer (DASViT), which bridges the gap in differentiable search for ViTs and uncovers novel designs and achieves superior efficiency with fewer parameters and FLOPs.
- 1
- 1
- Robust Long-Tailed Image Classification via Adversarial Feature Re-Calibration1
This paper presents Adversarial Feature Re-calibration (AFR), a method that improves the standard accuracy of a trained deep network by adding adversarial perturbations to the majority samples of each class by fine-tuning an adversarial attack model.
- 1
- 1
- Separable Batch Normalization for Robust Facial Landmark Localization1
A novel Separable Batch Normalization module with a Cross-protocol Network Training (CNT) strategy for robust facial landmark localization and a novel attention mechanism that assigns different weights to each branch for automatic selection in an effective style are presented.
- 1
- 1
- –
- Toward Stronger 3D Human Action Representation Learning Via Efficient Spatiotemporal Decoupling–
This work presents Stronger Spatiotemporal Decoupling Network (Stronger-SCD-Net), an effective self-supervised framework for 3D human action representation that systematically redesigns key components of the contrastive learning paradigm, including data augmentation, encoder architecture, and optimisation objective.
- Fastere: a fast framework for entity relation extractions–
Experimental results on benchmark datasets demonstrate that FastERE achieves speedup over state-of-the-art methods while maintaining competitive accuracy, with F1 scores of 69.9% and 67.0% on ACE05 and ACE04 respectively.
- Feature Alignment for Adversarial Robustness with Discriminative and Compact Representation Learning–
- Interactive image-to-video transfer learning–
An efficient image-to-video transfer learning framework, termed SDST (Static-Dynamic & Spatial-Temporal interaction framework), which explicitly enhances the interaction between static and dynamic cues as well as spatial and temporal domains is proposed, which surpasses State-of-the-Art efficient transfer learning techniques.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Robust Pedestrian Detection for Semi-automatic Construction of a Crowded Person Re-Identification Dataset–
A two-stage robust method for pedestrian detection in a complex crowded background to provide bounding box annotations and the identity of the detected subjects throughout each video is automatically annotated using multiple features and spatial-temporal clues.
- Contextual constraints based kernel discriminant analysis for face recognition–
This paper proposes a two-step subspace learning framework by combining non-linear kernel PCA (KPCA) and with contextual constraints based linear discriminant analysis (CCLDA) for face recognition, and Experimental results on three well-known face databases validate the effectiveness of the proposed method.
- –
- –
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.