Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileAcademic lineage
Possible advisorsa guess from early papers, not confirmed
- Rong ChenSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works189 from public data
- Bayesian Loss for Crowd Count Estimation With Point Supervision671
This work proposes Bayesian loss, a novel loss function which constructs a density contribution probability model from the point annotations, and outperforms previous best approaches by a large margin on the latest and largest UCF-QNRF dataset.
- Few-Shot Class-Incremental Learning587
This paper proposes the TOpology-Preserving knowledge InCrementer (TOPIC) framework, which mitigates the forgetting of the old classes by stabilizing NG's topology and improves the representation learning for few-shot new classes by growing and adapting NG to new training samples.
- Infrared-Visible Cross-Modal Person Re-Identification with an X Modality448
An X-Infrared-Visible (XIV) ReID cross-modal learning framework is proposed, which achieves an absolute gain of over 7% in terms of rank 1 and mAP even compared with the latest state-of-the-art methods.
- S-Prompts Learning with Pre-trained Transformers: An Occam's Razor for Domain Incremental Learning401
This paper proposes one simple paradigm (named as S-Prompting) and two concrete approaches to highly reduce the forgetting degree in one of the most typical continual learning scenarios, i.e., domain increment learning (DIL).
- Learning Graph Convolutional Network for Skeleton-Based Human Action Recognition by Neural Searching371
The first automatically designed GCN for skeleton-based action recognition is proposed and enriched by providing multiple dynamic graph modules after fully exploring the spatial-temporal correlations between nodes, and a sampling- and memory-efficient evolution strategy is proposed to search an optimal architecture for this task.
- Remote Heart Rate Measurement From Highly Compressed Facial Videos: An End-to-End Deep Learning Solution With Video Enhancement341
Comprehensive experiments are performed to show that the proposed method not only achieves superior performance on compressed videos with high-quality videos pair, but also generalizes well on novel data with only compressed videos available, which implies the promising potential for real-world applications.
- Spontaneous facial micro-expression analysis using Spatiotemporal Completed Local Quantized Patterns303
Experimental results demonstrate that STCLQP achieves a substantial improvement for analyzing facial micro-expressions, and an efficient vector quantization and discriminative codebook selection is proposed to make LBP-TOP more discrim inative and compact.
- Spatiotemporal Recurrent Convolutional Networks for Recognizing Spontaneous Micro-Expressions290
A novel deep recurrent convolutional networks based micro-expression recognition approach, capturing the spatiotemporal deformations of micro- expression sequence and optimized by an end-to-end manner and obviates manual feature design is proposed.
- Boosting Crowd Counting via Multifaceted Attention257
A Multifaceted Attention Network (MAN) is proposed to improve transformer models in local spatial relation encoding and an Instance Attention mechanism is provided to focus on the most important instances dynamically during training.
- Few-Shot Class-Incremental Learning via Relation Knowledge Distillation230
This paper proposes the exemplar relation distillation incremental learning framework to balance the tasks of old-knowledge preserving and new-knowledge adaptation, and outperforms other state-of-the-art class-incremental learning methods on the CIFAR100, miniImageNet, and CUB200 datasets.
- 213
- Topology-Preserving Class-Incremental Learning212
A novel topology-preserving class-incremental learning (TPCIL) framework that penalizes the changes of EHG’s neighboring relationships during incremental learning phases and demonstrates the power of the TPCIL for continuously learning new classes with less forgetting.
- A review of recent advances in visual speech decoding187
A detailed review of recent advances in visual speech decoding, focusing on the important questions asked by researchers and summarize the recent studies that attempt to answer them, and providing details of audio-visual speech databases.
- Image-to-image Translation via Hierarchical Style Disentanglement165
Hierarchical Style Disentanglement is proposed, which organizes the labels into a hierarchical tree structure, in which independent tags, exclusive attributes, and disentangled styles are allocated from top to bottom.
- 159
- Short and Long Range Relation Based Spatio-Temporal Transformer for Micro-Expression Recognition158
A comprehensive evaluation of the proposed spatio-temporal transformer architecture shows that the proposed approach consistently outperforms the state of the art, and is the first framework in the published literature on micro-expression recognition to achieve the unweighted F1-score greater than 0.9.
- 135
- Recurrent Convolutional Neural Network Regression for Continuous Pain Intensity Estimation in Video127
A real-time regression framework based on the recurrent convolutional neural network for automatic frame-level pain intensity estimation that can predict the pain intensity of each frame by considering a sufficiently large historical frames while limiting the scale of the parameters within the model.
- Universal Perturbation Attack Against Image Retrieval122
This paper proposes a novel method to generate retrieval-against UAP to break the neighbourhood relationships of image features via degrading the corresponding ranking metric, and proposes a multi-scale random resizing scheme and a ranking distillation strategy.
- 120
- Learning to Count via Unbalanced Optimal Transport113
This paper forms crowd counting as a measure regression problem to minimize the distance between two measures with different supports and unequal total mass, and adopts the unbalanced optimal transport distance to quantify the discrepancy between predicted density maps and point annotations.
- Dynamic texture and scene classification by transferring deep image features101
Inspired by the success of deep structures in image classification, a well-trained Convolutional Neural Network is applied as a feature extractor to extract mid-level features from each frame, and then form the video-level representation by concatenating the first and the second order statistics over the mid- level features.
- Isolated Sign Language Recognition with Grassmann Covariance Matrices94
This article proposes a covariance matrix--based representation to naturally fuse information from multimodal sources to utilize long-term dynamics over an isolated sign sequence, and demonstrates that the proposed method outperforms the state-of-the-art methods both in accuracy and computational cost.
- A Continual Deepfake Detection Benchmark: Dataset, Methods, and Essentials91
A continual deepfake detection benchmark (CDDB) over a new collection of deepfakes from both known and unknown generative models is suggested, which is clearly more challenging than the existing benchmarks, and offers a suitable evaluation avenue to the future research.
- 81
- Superpixel Masking and Inpainting for Self-Supervised Anomaly Detection79
A comprehensive evaluation of SMAI on the latest MVTec anomaly detection dataset shows that the unsupervised method plays favorably against state-of-the-art methods.
- Cross-Database Micro-Expression Recognition: A Benchmark78
A CDMER experimental evaluation protocol is established and a standard platform for evaluating their proposed methods is provided and extensive benchmark experiments are conducted by using NINE state-of-the-art domain adaptation (DA) methods and SIX popular spatiotemporal descriptors for investigating the CDMER problem from two different perspectives.
- 72
- 69
- 69
- 68
- Co-Attentive Lifting for Infrared-Visible Person Re-Identification64
A novel attention-based approach to handle the two difficulties in a unified framework of infrared-visible cross-modality person re-identification and proposes an attention lifting mechanism to learn discriminative features in each modality.
- 60
- 59
- 58
- Tripool: Graph triplet pooling for 3D skeleton-based action recognition57
Tripool is proposed, a novel graph pooling method for 3D action recognition from skeleton data that overcomes the limitation of the topology constrain for the high-level semantic representations, thus improves the performance.
- Micro-expression spotting: A new benchmark57
An extension of the SMIC-E database is introduced, namely the SMic-E-Long database, which is a new challenging benchmark for ME spotting and a new evaluation protocol is suggested that standardizes the comparison of various ME spotting techniques.
- Structured Modeling of Joint Deep Feature and Prediction Refinement for Salient Object Detection56
This paper designs a novel cascade CRFs architecture with CNN to jointly refine deep features and predictions at each scale and progressively compute a final refined saliency map, and formulate the CRF graphical model that involves message-passing of feature-feature, feature-prediction, and prediction-predicted, from the coarse scale to the finer scale.
- 56
- Direct Measure Matching for Crowd Counting54
This paper proposes a new measure-based counting approach to regress the predicted density maps to the scattered point-annotated ground truth directly, and derives a semi-balanced form of Sinkhorn divergence, based on which a S sinkhorn counting loss is designed for measure matching.
- 53
- 53
- Learning Scales from Points53
This paper proposes a density pyramid network (DPN), where each pyramid level handles instances within a particular scale range, and adopts an instance-level probabilistic scale-aware model (IPSM) to guide the multi-scale training of DPN explicitly.
- 48
- Semi-supervised Crowd Counting via Density Agency47
This paper builds a learnable auxiliary structure, namely the density agency to bring the recognized foreground regional features close to corresponding density sub-classes (agents) and push away background ones and proposes a density-guided contrastive learning loss to consolidate the backbone feature extractor.
- Reading Hidden Emotions: Spontaneous Micro-expression Spotting and Recognition.46
This work proposes the first automatic ME analysis system (MESR), which can spot and recognize MEs from spontaneous video data, and shows that this method achieves comparable performance to humans at this very challenging task, and outperforms humans in the ME recognition task by a large margin.
- 45
- 43
- A spatial-aware joint optic disc and cup segmentation method43
A spatial-aware joint segmentation method by explicitly considering the spatial locations of the pixels and learning the multi-scale spatially dense features is proposed, and high correlation between the cup-to-disk values and the risks of the glaucoma is validated on the dataset ORIGA.
- 43
- Age Estimation Using Local Binary Pattern Kernel Density Estimate43
The constructed age estimation system provides comparable performance against the state-of-the-art methods and a comparison of two local binary pattern based facial representations, namely a spatially enhanced histogram and a novel kernel density estimate.
- 42
- Anomaly Detection Via Self-Organizing Map41
The proposed method, Self-organizing Map for Anomaly Detection (SOMAD) maintains normal characteristics by using topological memory based on multi-scale features and achieves state-of-the-art performance on unsupervised anomaly detection and localization on the MVTec dataset.
- Transductive semi-supervised metric learning for person re-identification40
This paper focuses on the challenging semi-supervised person Re-identification (ReID) task, which is a metric learning problem based on the assumption that unlabeled data is open-set and proposes the Transductive Semi-Supervised Metric Learning (TSSML) framework.
- Video Action Recognition Via Neural Architecture Searching40
This paper makes the first attempt to let algorithm automatically design neural networks for video action recognition tasks using a spatio-temporal network developed in a differentiable space modeled by a directed acyclic graph.
- Capturing correlations of local features for image representation40
The superiority of the second-order standardized moment average pooling (2Standmap) is suggested and 2Standmap is successfully applied to four challenging tasks namely texture classification, medical image analysis, pain expression recognition, and micro-expression recognition.
- 38
- 37
- 36
- Error-Aware Density Isomorphism Reconstruction for Unsupervised Cross-Domain Crowd Counting35
A novel Error-aware Density Isomorphism REConstruction Network (EDIREC-Net) is proposed for cross-domain crowd counting that jointly transfers a pre-trained counting model to target domains using a density isomorphism reconstruction objective and models the reconstruction erroneousness by error reasoning.
- 35
- Improved Spatiotemporal Local Monogenic Binary Pattern for Emotion Recognition in The Wild34
An improved spatiotemporal feature descriptor based on STLMBP is proposed, which uses not only magnitude and orientation, but also the phase information, which provide complementary information in the robustness of LBP-TOP to unconstrained environment.
- Semi-Supervised Crowd Counting With Contextual Modeling: Facilitating Holistic Understanding of Crowd Scenes33
This paper presents a new semi-supervised method based on the mean teacher framework that achieves the state-of-the-art performance, surpassing previous approaches by a large margin on challenging benchmarks such as ShanghaiTech A and UCF-QNRF.
- Bi-Objective Continual Learning: Learning ‘New’ While Consolidating ‘Known’33
This paper develops the contrastive pillar (CPL) loss term to improve the classification performance, and examines several data sampling strategies for efficient onsite learning from ‘new’ with a reasonable amount of computational resources.
- 33
- SpikeMba: Multi-Modal Spiking Saliency Mamba for Temporal Video Grounding31
The SpikeMba approach integrates Spiking Neural Networks with state space models to leverage their unique advantages in handling different aspects of the task, and addresses confidence bias and long-term dependencies, thereby significantly enhancing fine-grained multimodal relationship capture.
- LBP-TOP: A Tensor Unfolding Revisit31
The results show that the fast LBP-TOP implementation is much more time-saving than the original one, and the concept of tensor unfolding method to compute three concatenated big matrices in two-dimensional space is introduced.
- Can SAM Count Anything? An Empirical Study on SAM Counting29
This study explores the use of SAM for the challenging task of few-shot object counting, which involves counting objects of an unseen category by providing a few bounding boxes of examples.
- 29
- OpenSDI: Spotting Diffusion-Generated Images in the Open World28
MaskCLIP is introduced, an SPM-based model that aligns Contrastive Language-Image Pre-Training (CLIP) with Masked Autoencoder with Masked Autoencoder (MAE) and significantly outperforms current state-of-the-art methods for the OpenSDI challenge.
- 28
- 27
- VisDrone-MOT2020: The Vision Meets Drone Multiple Object Tracking Challenge Results27
The VisDrone-MOT2020 consists of 79 challenging video sequences, including 56 videos for training, 7 videos for validation and 17 videos for evaluation, which feature 12 participating MOT algorithms.
- Dense prediction for micro-expression spotting based on deep sequence model27
This paper proposes a novel method for ME spotting based on a deep sequence model that uses a LSTM network that can utilize both local and global correlation of the extracted feature to predict the score of the ME apex frame.
- T2ICount: Enhancing Cross-Modal Understanding for Zero-Shot Counting26
A Hierarchical Semantic Correction Module that progressively refines text-image feature alignment, and a Representational Regional Coherence Loss that provides reliable supervision signals by leveraging the cross-attention maps extracted from the demising U-Net are proposed.
- K-armed Bandit based Multi-Modal Network Architecture Search for Visual Question Answering26
The experimental results show that with the help of KAB-NAS, AGAN can achieve the state-of-the-art performance on both benchmark datasets with much fewer parameters and computations.
- OMEG: Oulu Multi-Pose Eye Gaze Dataset26
Large datasets with well-defined gaze directions are desired for pattern recognition tasks including eye gaze estimation because most researchers used normal face datasets, which are not specifically designed forEye gaze estimation.
- Sliding Window Based Micro-expression Spotting: A Benchmark25
A benchmark for fairer and better performance evaluation of micro-expression spotting approaches is constructed and a sliding window based multi-scale evaluation standard with a series of protocols is proposed.
- ECCNAS: Efficient Crowd Counting Neural Architecture Search24
A novel search from pre-trained strategy enables a cross-task NAS to explore the significantly large and flexible search space with less search time and get more proper network structures, and develops a novel latency cost estimation algorithm in this ECCNAS.
- MEGC202224
This challenge focuses on two tasks, i.e., the micro- and macro-expression spotting task, and the ME Generation task.
- Texture Description with Completed Local Quantized Patterns22
The proposed completed local quantized patterns (CLQP) uses vector quantization to learn three separate codebooks for local sign, magnitude and orientation patterns, respectively and outperforms well-established features including LBP, LTP, CLBP,LQP on a range of challenging texture classification problems and an infant pain detection problem.
- One-Shot Replay: Boosting Incremental Object Detection via Retrospecting One Object20
A novel One-Shot Replay (OSR) method for incremental object detection, which is an augmentation-based method, rather than storing original images, only one object-level sample for each old class is stored to reduce memory usage significantly, and it is found that copy-paste is a harmonious way to replay for incrementalobject detection.
- FME'2120
This workshop will explore various dimensions of the human mind through emotion understanding and FME analysis, as well as extended research based on multi modal approaches.
- Facial Action Unit Detection with Local Key Facial Sub-region based Multi-label Classification for Micro-expression Analysis19
A segmentation method for key facial sub-regions based on the location of AUs and facial landmarks, which extracts 11 facial key regions from each sequence of micro-expression images and is shown to exhibit state-of-the-art performance.
- Image Captioning via Dynamic Path Customization18
This article explores a novel dynamic network for vision and language (V&L) tasks, where the inferring structure is customized on the fly for different inputs, to build a rich routing space and improve routing efficiency.
- 18
- Know Your Surroundings: Panoramic Multi-Object Tracking by Multimodality Collaboration18
A MultiModality PAnoramic multi-object Tracking framework (MMPAT), which takes both 2D panorama images and 3D point clouds as input and then infers target trajectories using the multimodality data.
- Cross-database Micro-Expression Recognition with Deep Convolutional Networks18
In this paper, an end-to-end deep model is employed for learning the representation and classifier automatically and the recurrent convolutional layers are utilized to exploit the learning ability with the optical flow fields of micro-expression sequences, which are enhanced by a motion magnification procedure.
- 17
- 17
- 17
- 17
- Structured partial least squares for simultaneous object tracking and segmentation17
This paper proposes a novel method to address this limitation by incorporating structured labeling information in the partial least square analysis algorithms for simultaneous object tracking and segmentation, which allows for novel structured labeling constraints to be placed directly on the tracked objects to provide useful contour constraint to alleviate the drifting problem.
- Semi-Supervised Counting via Pixel-by-Pixel Density Distribution Modeling16
This paper designs a pixel-wise distribution matching loss to measure the differences in the pixel-wise density distributions between the prediction and the ground-truth, and designs the interleaving consistency self-supervised learning mechanism to learn from unlabeled data efficiently.
- 16
- 16
- Pseudo Object Replay and Mining for Incremental Object Detection16
This paper proposes a pseudo object replay and mining method (PseudoRM) to handle the co-occurrence dependent problem, reducing the performance degradation caused by the absence of old-class objects.
- Benchmarking Deepart Detection16
This paper establishes a deepart detection database (DDDB) that consists of a set of high-quality conventional art images (conarts) and five sets of deepart images generated by five state-of-the-art deepfake models, and suggests four benchmark evaluations and four families of solutions on the constructed DDDB.
- Deep Class-Incremental Learning From Decentralized Data16
This article introduces a paradigm to create a basic decentralized counterpart of typical (centralized) CIL approaches, and proposes a decentralized composite knowledge incremental distillation framework to transfer knowledge from historical models and multiple local sites to the general model continually.
- Boosting Semi-supervised Crowd Counting with Scale-based Active Learning15
This paper proposes a simple yet effective active labeling strategy to explicitly select informative unlabeled images, guided by the intra-scale uncertainty and inter-scale inconsistency metrics, and implements a progressive updating scheme for the semi-supervised crowd counting framework.
- Saliency Integration: An Arbitrator Model15
This paper proposes an arbitrator model (AM) for saliency integration that incorporates the consensus of multiple saliency models and the external knowledge into a reference map to effectively rectify the misleading by candidate models and derives a Bayesian integration framework to reconcile theSaliency models of varying expertise and the reference map.
- Specifying What You Know or Not for Multi-Label Class-Incremental Learning14
This paper targets at specifying what is known or not to accommodate Historical, Current, and Prospective knowledge for MLCIL and proposes a novel framework termed as HCP, which clarifies the known classes by dynamic feature purification and recall enhancement with distribution prior, enhancing the precision and retention of known information.
- 14
- 13
- MEGC2023: ACM Multimedia 2023 ME Grand Challenge12
This challenge focuses on facial micro-expressions and the macro-expression (MaE) spotting task and builds an unseen cross-cultural long-video test set based on CAS(ME)2, SAMM Long Videos, SMIC-E-long, CAS( ME)3 and 4DME.
- 12
- 12
- PCANet-II: When PCANet Meets the Second Order Pooling12
Compared with the histogram-based output, the second order pooling not only provides more discriminative information by preserving both the magnitude and sign of convolutional responses, but also dramatically reduces the size of output features.
- HIT-AVDB-II: A New Multi-view and Extreme Feature Cases Contained Audio-Visual Database for Biometrics12
A multi-language and multiview database HIT-AVDB-II with a corpus of various common and special sentences include Chinese and English poems, tongue twister, digits, Greek alphabet and music is constructed.
- DCA: Dividing and Conquering Amnesia in Incremental Object Detection11
This paper proposes a Divide-and-Conquer Amnesia (DCA) strategy, which redesigns the transformer-based IOD into a localization-then-recognition process, which can well maintain and transfer the localization ability, leaving decoupled fragile recognition to be specially conquered.
- Sparse projections matrix binary descriptors for face recognition11
This paper introduces the sparsity constraint into the objective function of general binary codes learning framework, so that the problem of high computational cost and overfitting can be somehow solved and learns a discriminative and low-cost Sparse Projection Matrix Binary Descriptors (SPMBD).
- 11
- 10
- A Part Power Set Model for Scale-Free Person Retrieval10
An end-to-end part power set model with multi-scale features is proposed, which captures the discriminative parts of pedestrians from global to local, and from coarse to fine, enabling part-based scale-free person re-ID.
- A unified 3D face authentication framework based on robust local mesh SIFT feature10
A facial depth recovery method to construct a facial depth map from stereoscopic videos and a novel feature descriptor is proposed, called Local Mesh Scale-Invariant Feature Transform (LMSIFT) to reflect the different face recognition abilities in different facial regions.
- 9
- Multi-modal Crowd Counting via Modal Emulation9
A modal emulation-based two-pass multi-modal crowd-counting framework that enables efficient modal emulation, alignment, and fusion, and introduces a modality alignment module that uses an efficient modal consistency loss to align the outputs of the two passes and bridge the semantic gap between modalities.
- 9
- 9
- Characterizing Subtle Facial Movements via Riemannian Manifold9
The tangent delta descriptor is proposed to characterize the dynamics of short-term movements using the differences between points on the tangent spaces to the manifolds, rather than the points themselves, to diminish the influences of extrinsic diversity.
- Bidirectional Long Short-Term Memory Variational Autoencoder.9
The proposed bLSTM-VAE, a novel variant of VAE whose encoders and decoders are implemented by bidirectional Long Short-Term Memory (bLSTM) networks, could learn latent variables reliably by mining the contextual information from the whole input sequence.
- 8
- 8
- MEGC2025: Micro-Expression Grand Challenge on Spot Then Recognize and Visual Question Answering7
Two tasks that reflect evolving research directions in facial micro-expressions are introduced: Me spot-then-recognize (ME-STR), which integrates ME spotting and subsequent recognition in a unified sequential pipeline; and ME visual question answering (ME-VQA), which explores ME understanding through visual question answering, leveraging MLLMs or LVLMs to address diverse question types related to MEs.
- 7
- MEGC2024: ACM Multimedia 2024 Facial Micro-Expression Grand Challenge7
This Grand Challenge comprises two distinct tracks: the Cross-Cultural Spotting (CCS) track, and the Spot-Then-Recognize (STR) track; all participating solutions submitted their results to a leaderboard, and several submissions performed well surpassing their respective baseline results.
- 7
- Universal Adversarial Perturbations Against Person Re-Identification.7
This paper attacks the re-ID models using universal adversarial perturbations (UAPs), which are especially dangerous to the surveillance systems because it could fool most pedestrian images with a little overhead, and proposes an effective method to train UAPs against person re-IDs from the global list-wise perspective.
- 7
- 7
- 7
- Weighted Non-locally Self-similarity Sparse Representation for Face Deblurring7
This paper introduces the current weighted non-locally self-similarity (WNLSS) method, which is originally proposed to remove the noise for natural images, into the face deblurring model and shows the properties of data fidelity term and regularization term also can fit well for facedeblurring problem.
- Constructing Local Binary Pattern Statistics by Soft Voting7
A novel method for constructing Local Binary Pattern statistics for image appearance description inspired by the kernel density estimation designed for estimating the underlying probability function of a random variable, which yields a more reliable estimate of the underlying LBP distribution of the given image.
- 6
- 6
- Saliency detection via bi-directional propagation6
The bi-directional propagation model (BIP) for saliency detection outperforms fourteen state-of-the-art saliency models on four challenging datasets, and largely enhances the propagation efficiency compared to single directional propagation models.
- Texture and Motion Pattern Fusion for Background Subtraction6
A novel background subtraction algorithm, which takes both texture and motion information into account, and combines the texture pattern- based and motion pattern-based background model.
- 5
- 5
- Few-Shot Audio-Visual Class-Incremental Learning with Temporal Prompting and Regularization5
An efficient framework based on Vision Transformer is proposed to solve FS-AVCIL and temporal-residual prompting is incorporated into the audio-visual adapter to efficiently finetune the pre-trained foundation model with limited data and capture audio-visual correlation by learning temporal-relevant prompts.
- 5
- 5
- 5
- A Task-Driven Eye Tracking Dataset for Visual Attention Analysis5
Inspired by psychological findings that human visual behavior is tightly dependent on the executed tasks, this work carefully design specific tasks in accordance with the contents of 111 images covering various semantic categories, such as text, facial expression, texture, pose, and gaze.
- Agent-MER: A Cognitive Agent with Hierarchical Deliberation for Open-Vocabulary Multimodal Emotion Recognition4
A novel cognitive agent framework (Agent-MER) is proposed, which reframes the OV-MER task as a problem to be solved by an agent that mimics the human cognitive process through knowledge-guided deliberation, enabling the identification of a richer and deeper range of emotions.
- Towards Practical Multi-Robot Hybrid Tasks Allocation for Autonomous Cleaning4
This paper proposes a novel robust mixed-integer linear programming model with practical constraints including the task order constraint for different tasks and the ability constraints of hybrid robots and provides comprehensive results on the collected dataset.
- Remind of the Past: Incremental Learning with Analogical Prompts4
This paper designs an analogy-making mechanism to remap the new data into the old class by prompt tuning, which consistently outperforms data-replay methods by only saving feature prototypes for each class.
- Structural Knowledge Organization and Transfer for Class-Incremental Learning4
A structural graph knowledge distillation based incremental learning framework to preserve both the positions of samples and their relations and develop a graph interpolation mechanism to enrich the domain of knowledge and alleviate the inter-class sample imbalance issue.
- An Information Acquiring Channel —— Lip Movement4
It could be easily got that lip-reading channel is an effective one and can be applied independently and the reasoning is given by describing two kinds of valid applications, which are constructed on lip movement information only.
- ComprehendEdit: A Comprehensive Dataset and Evaluation Framework for Multimodal Knowledge Editing3
This study introduces ComprehendEdit, a comprehensive benchmark comprising eight diverse tasks from multiple datasets and proposes two novel metrics: Knowledge Generalization Index (KGI) and Knowledge Preservation Index (KPI), which evaluate editing effects on in-domain samples without relying on AI-synthetic samples.
- 3
- 3
- FME '223
FME'22 is the inaugural workshop in micro-expression research, with the aim of promoting interactions between researchers and scholars from within this niche area of research and also including those from broader, general areas of expression and psychology research.
- Few-shot Learning for Multi-Modality Tasks3
To simulate the learning schema of humans and relieve the reliance on the large-scale annotation benchmarks, researchers start shifting towards the few-shot learning problem: they try to learn a model to correctly recognize novel categories with only a few annotated samples.
- 3
- An immersive fire training system using kinect3
Experimental results show the fire escape skills of 100 children are greatly improved and a simulation training system to help children learn fire hazards knowledge and escape skills is presented.
- 2
- Multimodal Blockwise Transformer for Robust Sentiment Recognition2
A novel multimodal Blockwise Transformer architecture is proposed, which effectively integrates visual, auditory, and textual features to improve emotion classification accuracy and demonstrates the effectiveness and robustness of this approach in handling noisy data for emotion recognition tasks.
- 2
- Salient Object Detection with CNNs and Multi-scale CRFs2
A more flexible CRF refinement framework by embedding the CRF inference to multiple levels of side outputs from CNNs for multi-scale saliency refinement and shows competitive performance over four datasets in comparison with the existing state-of-the-art saliency models.
- Spatial-temporal texture features for 3D human activity recognition using laser-based RGB-D videos2
A novel feature extraction framework to describe human activities based on the above optical video capturing method, namely spatial-temporal texture features for 3D human activity recognition, which is insensitive to illumination and occlusions, and efficient for fine-motion description.
- 2
- Sample-Aware Knowledge Association and Enhancement for Open-Vocabulary Continual Learning1
An Isolation-Association-Enhancement (IAE) framework is proposed that isolates task knowledge through modular lightweight parameters, dynamically associates relevant knowledge components based on sample-specific visual-semantic relationships, and integrates them via a Bayesian-driven mixture-of-experts strategy for robust, uncertainty-aware adaptation.
- 1
- 1
- Penny-Wise and Pound-Foolish in Deepfake Detection1
A novel learning framework (PoundNet) is proposed for generalization of deepfake detection on a pre-trained vision-language model that incorporates a learnable prompt design and a balanced objective to preserve broad knowledge from upstream tasks (object classification) while enhancing generalization for downstream tasks (deepfake detection).
- FME '23: 3rd Facial Micro-Expression Workshop1
FME'23 is the inaugural workshop in micro-expressions, with the aim of promoting interactions between researchers and scholars from within this niche area of research, and discussing the growing ethical conversations when using face data, and how to come to a consensus on micro-expression standards within affective computing.
- 1
- Higher order partial least squares for object tracking: A 4D-tracking method1
This paper proposes a novel 4D object tracking method via the higher order partial least squares (HOPLS) which is a generalized multi-linear regression method and uses HOPLS to adaptively learn low-dimensional discriminative feature subspace for object representation.
- Efficient Boosted Weak Classifiers for Object Detection1
Experimental results in the detection tasks for multiple classes of objects show that boosted PVCs significantly improves both learning and evaluation efficiency of nonlinear SVMs to the level of boosted linear classifiers, without losing any of the high discriminative power.
- 1
- 1
- Lipreading HLM and Text Flow Analysis1
A novel recognition model HLM (HMM and Bigram Language Model), which is based on HMM, and combined with prior knowledge of language is proposed, which combines closely language background knowledge and HMM.
- –
- –
- –
- –
- –
- A Benchmark for Incremental Micro-expression Recognition–
This work introduces the first benchmark specifically designed for incremental micro-expression recognition, and formulates the incremental learning setting tailored for micro-expression recognition.
- –
- Label-Efficient Emotion and Sentiment Analysis–
This tutorial focuses on the label-efficient ESA (LeESA) learning methods and introduces the stimuli and characteristics of emotion and illustrates seven typical training paradigms, followed by applications and future directions of LeESA.
- –
- –
- –
- –
- –
- –
- Research of pixel based feature extraction in lip-reading–
A three-stage cascade visual front end is proposed and PCA is applied to reduce the dimension of DCT and Gabor transformed data to improve the recognition accuracy by 10% compared with the manually-selected features.
- –
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.