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
- Yu QiaoSuggested from co-authorship
- Qiang PengSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works136 from public data
- Region Attention Networks for Pose and Occlusion Robust Facial Expression Recognition933
A novel Region Attention Network (RAN), to adaptively capture the importance of facial regions for occlusion and pose variant FER, and a region biased loss to encourage high attention weights for the most important regions.
- Suppressing Uncertainties for Large-Scale Facial Expression Recognition724
Self-Cure Network suppresses the uncertainty from two different aspects: 1) a self-attention mechanism over FER dataset to weight each sample in training with a ranking regularization, and 2) a careful relabeling mechanism to modify the labels of these samples in the lowest-ranked group.
- Bag of visual words and fusion methods for action recognition: Comprehensive study and good practice653
A comprehensive study of all steps in BoVW and different fusion methods is provided, and a simple yet effective representation is proposed, called hybrid supervector, by exploring the complementarity of different BoVW frameworks with improved dense trajectories.
- Multi-region Two-Stream R-CNN for Action Detection395
A multi-region two-stream R-CNN model for action detection in realistic videos that links frame-level detections with the Viterbi algorithm, and temporally localize an action with the maximum subarray method is proposed.
- Action Recognition with Stacked Fisher Vectors388
Experimental results demonstrate the effectiveness of SFV, and the combination of the traditional FV and SFV outperforms state-of-the-art methods on these datasets with a large margin.
- Emotion-LLaMA: Multimodal Emotion Recognition and Reasoning with Instruction Tuning298
The MERR dataset is introduced, containing 28,618 coarse-grained and 4,487 fine-grained annotated samples across diverse emotional categories and Emotion-LLaMA, a model that seamlessly integrates audio, visual, and textual inputs through emotion-specific encoders is proposed, significantly enhances both emotional recognition and reasoning capabilities.
- Attention-Driven Dynamic Graph Convolutional Network for Multi-label Image Recognition265
An Attention-Driven Dynamic Graph Convolutional Network (ADD-GCN) to dynamically generate a specific graph for each image to model the relation of content-aware category representations that are generated by a Semantic Attention Module (SAM).
- Visual Compositional Learning for Human-Object Interaction Detection239
A deep Visual Compositional Learning (VCL) framework is devised, which is a simple yet efficient framework to effectively address the problem of human-Object interaction detection and largely alleviates the long-tail distribution problem and benefits low-shot or zero-shot HOI detection.
- AnoPCN227
AnoPCN unifies reconstruction and prediction methods in an end-to-end framework, and it achieves state-of-the-art performance with better prediction results and larger regularity score gaps on three benchmark datasets including ShanghaiTech Campus, CUHK Avenue, and UCSD Ped2.
- 220
- Frame Attention Networks for Facial Expression Recognition in Videos212
This paper proposes the Frame Attention Networks (FAN)1, to automatically highlight some discriminative frames in an end-to-end framework and shows superior performance compared to other CNN based methods and achieves state-of-the-art performance on CK+.
- Affordance Transfer Learning for Human-Object Interaction Detection144
An affordance transfer learning approach is introduced to jointly detect HOIs with novel object and recognize affordances, and is capable of inferring the affordances of novel objects from known affordance representations.
- Detecting Human-Object Interaction via Fabricated Compositional Learning139
A novel HOI compositional learning framework is devised, termed as Fabricated Compositional Learning (FCL), to address the problem of open long-tailed HOI detection, which introduces an object fabricator to generate effective object representations, and then combines verbs and fabricated objects to compose new HOI samples.
- Frankenstein: Learning Deep Face Representations Using Small Data136
This paper proposes a method to generate very large training data sets of synthetic images by compositing real face images in a given data set and shows that this method enables to learn models from as few as 10 000 training images, which perform on par with models trained from 500 000 images.
- From Large Language Models to Large Multimodal Models: A Literature Review114
This paper provides a summary of both LLMs and LMMs from a unified perspective, makes an analysis of the development status of large-scale models in the view of globalization, and offers potential research directions for large-scale models.
- AffectGPT: A New Dataset, Model, and Benchmark for Emotion Understanding with Multimodal Large Language Models91
This work establishes a new benchmark for MLLM-based emotion understanding with a novel dataset (MER-Caption) and a new model (AffectGPT), and introduces the AffectGPT model, designed with pre-fusion operations to enhance multimodal integration.
- 91
- 79
- Deep Recurrent Multi-instance Learning with Spatio-temporal Features for Engagement Intensity Prediction78
This paper elaborates the winner approach for engagement intensity prediction in the EmotiW Challenge 2018, and builds a multi-modal regression model based on multi-instance mechanism as well as LSTM.
- Exploring Emotion Features and Fusion Strategies for Audio-Video Emotion Recognition74
The approaches in EmotiW 2019, which mainly explores emotion features and feature fusion strategies for audio and visual modality, are described and several facial features with different CNN models and different emotion pretrained strategies are evaluated.
- Group emotion recognition with individual facial emotion CNNs and global image based CNNs74
This paper presents the approach for group-level emotion recognition in the Emotion Recognition in the Wild Challenge 2017, based on two types of Convolutional Neural Networks, namely individual facial emotion CNNs and global image based CNNs.
- Boosting VLAD with Supervised Dictionary Learning and High-Order Statistics69
A high-order VLAD (H-VLAD) for visual recognition is proposed, which leverages two kinds of high- order statistics in the VL AD-like framework, namely diagonal covariance and skewness and allows for efficient computation.
- Cascade Attention Networks For Group Emotion Recognition with Face, Body and Image Cues46
This paper presents the approach for group-level emotion recognition sub-challenge in the EmotiW 2018 and proposes a cascade attention network for the face cue in images to generate a global representation based on all faces.
- Action and Gesture Temporal Spotting with Super Vector Representation46
This paper proposes an action and gesture spotting system, which is mainly composed of three steps: temporal segmentation, clip classification, and post processing, and uses some post processing techniques to void the detection of false positives.
- Exploring Motion Boundary based Sampling and Spatial-Temporal Context Descriptors for Action Recognition45
A motion boundary based dense sampling strategy is introduced, which greatly reduces the number of valid trajectories while preserves the discriminative power, and a set of new descriptors which describe the spatial-temporal context of motion trajectories are developed.
- Artificial-Intelligence-Generated Content with Diffusion Models: A Literature Review44
This survey aims to offer an exhaustive overview of both the theoretical underpinnings and practical achievements of diffusion models, exploring and outlining three core approaches to diffusion modeling: denoising diffusion probabilistic models, score-based generative models, and stochastic differential equations.
- Learning label correlations for multi-label image recognition with graph networks44
This paper introduces a plug-and-play Label Graph (LG) module to learn label correlations with word embeddings, and utilizes traditional GCN to map this graph into label-dependent object classifiers which are further applied to image features.
- Residual Compensation Networks for Heterogeneous Face Recognition44
This paper proposes a new two-branch network architecture, termed as Residual Compensation Networks (RCN), to learn separated features for different modalities in HFR, which outperforms other state-of-the-art methods significantly.
- CaT-GNN: Enhancing Credit Card Fraud Detection via Causal Temporal Graph Neural Networks42
A novel method for credit card fraud detection, the CaT-GNN, which leverages causal invariant learning to reveal inherent correlations within transaction data and applies a causal mixup strategy to enhance the model's robustness and interpretability.
- Suppressing Mislabeled Data via Grouping and Self-attention40
This paper proposes a conceptually simple yet efficient training block, termed as Attentive Feature Mixup (AFM), which allows paying more attention to clean samples and less to mislabeled ones via sample interactions in small groups, which has several appealing benefits for noise-robust deep learning.
- 38
- Unsupervised person re-identification with multi-label learning guided self-paced clustering36
Extensive experiments on three popular large-scale Re-ID benchmarks demonstrate that the MLC outperforms previous state-of-the-art methods and significantly improves the performance of unsupervised person Re- ID.
- An Efficient Training Approach for Very Large Scale Face Recognition35
This work proposes a novel training approach, termed Faster Face Classification (F2C), to alleviate time and cost without sacrificing the performance, and proves that the speed is faster than state-of-the-art FC-based methods in terms of recognition accuracy and hardware costs.
- A Closer Look at Time Steps is Worthy of Triple Speed-Up for Diffusion Model Training34
This paper introduces a novel speed-up method for diffusion model training, called SpeeD, which is based on a closer look at time steps, and proposes a weighting strategy to emphasize the importance of time steps with rapid-change process increments.
- Rail Detection: An Efficient Row-based Network and a New Benchmark34
A rail database and a row-based rail detection method, Rail-Net, containing a lightweight convolutional backbone and an anchor classifier, which achieves promising results in terms of both speed and accuracy.
- Cascade multi-head attention networks for action recognition33
A new attention network architecture, termed as Cascade multi-head ATtention Network (CATNet), which constructs video representations with two-level attentions, namely multi- head local self-attentions and relation based global attentions is proposed.
- 31
- A Challenge Dataset and Effective Models for Conversational Stance Detection30
This paper introduces a new multi-turn conversation stance detection dataset (called MT-CSD), which encompasses multiple targets for conversational stance detection, and proposes a global-local attention network (GLAN) to address both long and short-range dependencies inherent in conversational data.
- Semi-Supervised Multimodal Emotion Recognition with Expression MAE30
This paper presents the submission approach on the Semi-Supervised Learning Sub-Challenge (MER-SEMI), and with large-scale unlabeled emotional videos, trains both image-based and video-based Masked Autoencoders to extract visual features, which are termed as expression MAE (expMAE) for simplicity.
- Large Language Model Enhanced Logic Tensor Network for Stance Detection29
The FOLAR framework is introduced, an innovative approach that integrates first-order logic (FOL) with large language models (LLMs) to enhance the interpretability and efficacy of stance detection and is shown as a promising solution for explainable and accurate stance detection.
- Product image recognition with guidance learning and noisy supervision28
A novel large-scale product image dataset, termed as Product-90, and a simple yet efficient guidance learning method for training convolutional neural networks (CNNs) with noisy supervision, which achieves performance superior to state-of-the-art methods on these datasets.
- 26
- Bootstrap Model Ensemble and Rank Loss for Engagement Intensity Regression25
This paper presents the approach for the engagement intensity regression task of EmotiW 2019, and uses the classical bootstrap aggregation method to perform model ensemble which randomly samples a certain training data by several times and then averages the model predictions.
- Invisible gas detection: An RGB-thermal cross attention network and a new benchmark24
The RGB-Thermal Cross Attention Network (RT-CAN) is presented, which employs an RGB-assisted two-stream network architecture to integrate texture information from RGB images and gas area information from thermal images, achieving state-of-the-art (SOTA) performance among RGB-thermal methods.
- Image Encryption with The Cross Diffusion of Two Chaotic Maps24
This paper proposes an image encryption method based on the cross diffusion of two chaotic maps which has the characteristics of large key space, fast, robust to statistic attack, etc.
- TTPP: Temporal Transformer with Progressive Prediction for efficient action anticipation22
A simple yet efficient Temporal Transformer with Progressive Prediction (TTPP) framework is proposed, which repurposes a Transformer-style architecture to aggregate observed features, and then leverages a light-weight network to progressively predict future features and actions.
- 22
- Motion boundary based sampling and 3D co-occurrence descriptors for action recognition22
A motion boundary based sampling strategy and spatial-temporal co-occurrence descriptors for action video representation and recognition and the results show that the sampling strategy significantly reduces the computational cost of point tracking without degrading performance.
- Leveraging Chain-of-Thought to Enhance Stance Detection with Prompt-Tuning20
A novel prompt-based stance detection method that leverages the knowledge acquired using the chain-of-thought method, which is referred to as PSDCOT and achieves state-of-the-art results in in-domain, cross-target, and zero-shot learning settings.
- Hybrid Super Vector with Improved Dense Trajectories for Action Recognition20
This work employs two advanced super vector methods, namely Fisher Vector and soft Vector of Locally Aggregated Descriptors to encode them separately, and concatenates them into a Hybrid Super Vector, and a linear SVM classifier is used to predict labels.
- Joint 3D facial shape reconstruction and texture completion from a single image19
The qualitative and quantitative results demonstrate the effectiveness of the SRTC-Net methods on inferring 3D facial geometry and complete texture; they outperform or are comparable to the state-of-the-art.
- 19
- Exploring Regularizations with Face, Body and Image Cues for Group Cohesion Prediction18
This paper elaborately design two regularizations, namely a rank loss and a hourglass loss, where the former aims to give a margin between the distance of distant categories and near categories and the later to avoid centralization predictions with only MSE loss.
- Encoding Feature Maps of CNNs for Action Recognition18
This work utilizes a recent deep CNN model, VGG19, to capture appearance features and uses VLAD encoding to encode/pool convolutional feature maps which shows better performance than average pooling of feature maps and full-connected activation features.
- IncepFormer: Efficient Inception Transformer with Pyramid Pooling for Semantic Segmentation17
IncepFormer introduces a novel pyramid structured Transformer encoder which harvests global context and fine localisation features simultaneously simultaneously, and integrates an Inception-like architecture with depth-wise convolutions, and a light-weight feed-forward module in each self-attention layer, efficiently obtaining rich local multi-scale object features.
- 17
- 17
- Recurrent Metric Networks and Batch Multiple Hypothesis for Multi-Object Tracking17
This paper proposes the recurrent metric network (RMNet), a convolutional neural network-recurrent neuralnetwork-based similarity metric framework for the multi-object tracking, and introduces a batch multiple hypothesis (BMH) strategy, a simple yet efficient data association method for the batch multi- object tracking.
- Sketch-based 3-D modeling for piecewise planar objects in single images16
Experimental results on various objects show that the proposed approach provides a good solution to these three problems: requiring rough sketch input only, accurate modeling for a large class of objects, and automatically recovering the invisible part of an object and providing a complete 3-D model.
- miMamba: EEG-Based Emotion Recognition With Multi-Scale Inverted Mamba Models14
A novel network called Multi-scale Inverted Mamba, which consists of Multi-Scale Temporal Blocks (MSTB) and Temporal-Spatial Fusion Blocks (TSFB), designed to capture both local details and global temporal dependencies across different scale subsequences, which underscores the model’s superior performance in multidimensional emotion recognition tasks and its potential for practical applications in resource-constrained affective computing scenarios.
- Exploring dense trajectory feature and encoding methods for human interaction recognition14
A finding that the performance of vector quantization is on par with or even better than other sophisticated feature encoding methods by using dense trajectories in videos is thoroughly analysed.
- Two in One Go: Single-stage Emotion Recognition with Decoupled Subject-context Transformer13
This work presents a single-stage emotion recognition approach, employing a Decoupled Subject-Context Transformer (DSCT), for simultaneous subject localization and emotion classification, and introduces DSCT to facilitate interactions between fine-grained subject-context cues in a ''decouple-then-fuse'' manner.
- 13
- 13
- Why We Feel: Breaking Boundaries in Emotional Reasoning with Multimodal Large Language Models12
This work presents EIBench, a large-scale benchmark encompassing 1615 basic EI samples and 50 complex EI samples featuring multifaceted emotions, and proposes a Coarse-to-Fine Self-Ask (CFSA) annotation pipeline, which guides Vision-Language Models (VLLMs) through iterative question-answer rounds to yield high-quality labels at scale.
- Distantly Supervised Explainable Stance Detection via Chain-of-Thought Supervision12
A distantly supervised explainable stance detection framework (DS-ESD), comprising an instruction-based chain-of-thought method, a generative network, and a transformer-based stance predictor, and a label rectification strategy to mitigate the impact of erroneous labels is proposed.
- 12
- The Snowflake Hypothesis: Training Deep GNN with One Node One Receptive field12
The Snowflake Hypothesis is introduced -- a novel paradigm underpinning the concept of ``one node, one receptive field'' that can serve as a universal operator for a range of tasks, and it displays tremendous potential on deep GNNs.
- A Comprehensive Interaction in Multiscale Multichannel EEG Signals for Emotion Recognition11
A cascade scale-aware adaptive graph convolutional network and cross-EEG transformer (SAG-CET) is proposed to explore the comprehensive interaction between multiscale and multichannel EEG signals with two novel ideas.
- AU-Guided Unsupervised Domain-Adaptive Facial Expression Recognition11
An AU-guided unsupervised domain-adaptive FER (AdaFER) framework to relieve the annotation bias between different FER datasets and shows that AdaFER achieves state-of-the-art results on all these benchmarks.
- UMETTS: A Unified Framework for Emotional Text-to-Speech Synthesis with Multimodal Prompts10
The UMETTS framework is introduced, a novel framework that leverages emotional cues from multiple modalities to generate highly expressive and emotionally resonant speech and achieves significant improvements in emotion accuracy and speech naturalness, outperforming traditional E-TTS methods on both objective and subjective metrics.
- 10
- SZTU-CMU at MER2024: Improving Emotion-LLaMA with Conv-Attention for Multimodal Emotion Recognition9
This system leverages the advanced emotional understanding capabilities of Emotion-LLaMA to generate high-quality annotations for unlabeled samples, addressing the challenge of limited labeled data, and introduces Conv-Attention, a lightweight and efficient hybrid framework to enhance multimodal fusion while mitigating modality-specific noise.
- EA-Vit: Efficient Adaptation for Elastic Vision Transformer8
An efficient ViT adaptation framework is proposed that enables a single adaptation process to generate multiple models of varying sizes for deployment on platforms with various resource constraints and adopts a curriculum-based training strategy that progressively increases elasticity.
- Dynamic Vision Mamba8
The proposed method, Dynamic Vision Mamba (DyVM), effectively reduces FLOPs with minor performance drops and generalizes well across different Mamba vision model architectures and different vision tasks.
- The Snowflake Hypothesis: Training and Powering GNN with One Node One Receptive Field8
The Snowflake Hypothesis -- a novel paradigm underpinning the concept of "one node, one receptive field'' is introduced, proposing a corresponding uniqueness in the receptive fields of nodes in the GNNs.
- Gradient‐based compressive sensing for noise image and video reconstruction8
A fast gradient-based compressive sensing (FGB-CS) for noise image and video is proposed, and the authors formulate the convex optimisation of noise signal reconstruction under Lipschitz gradient and replace the iteration parameter by the LipsChitz constant.
- sEMG-Based Joint Angle Estimation via Hierarchical Spiking Attentional Feature Decomposition Network7
- An empirical study on temporal modeling for online action detection7
This paper aims to provide an empirical study on temporal modeling for OAD including four meta types of temporal modeling methods, i.e. temporal pooling, temporal convolution, recurrent neural networks, and temporal attention, and uncover some good practices to produce a state-of-the-art OAD system.
- Rethinking Structure Learning For Graph Neural Networks6
A new GSL framework is proposed, which includes three steps: GSL base (the representation used for GSL) construction, new structure construction, and view fusion, to better understand the effectiveness of GSL in GNNs, and it is found that it is the pretrained GSL bases that enhance GNN performance, and in most cases, GSL cannot improve GNN performance.
- 6
- 5
- Faster Vision Mamba is Rebuilt in Minutes via Merged Token Re-training5
This work shows how simple and effective the fast recovery can be achieved at minute-level, in particular, a 35.9% accuracy spike over 3 epochs of training on Vim-Ti, and that a quick round of retraining after token merging yeilds robust results across various compression ratios.
- 5
- FlexEdit: Marrying Free-Shape Masks to VLLM for Flexible Image Editing5
FlexEdit is proposed, an end-to-end image editing method that leverages both free-shape masks and language instructions for Flexible Editing and achieves state-of-the-art (SOTA) performance in LLM-based image editing.
- 5
- AU-Aware Vision Transformers for Biased Facial Expression Recognition5
This paper proposes a simple yet conceptually-new framework, AU-aware Vision Transformer (AU-ViT), which improves the performance of individual datasets by jointly training auxiliary datasets with AU or pseudo-AU labels and proves that a carefully-initialized ViT achieves comparable performance to advanced deep convolutional networks.
- 5
- 5
- Finding hard faces with better proposals and classifier4
Experimental results show that the boosted faster RCNN version with an enhanced region proposal network (eRPN) module and newly introduced hard example mining strategies significantly improves the faster-RCNN face detector and achieves performance superior or comparable to the state-of-the-art face detectors.
- An example-based approach to 3D man-made object reconstruction from line drawings4
An example-based approach to reconstructing 3D object, either planar or curved, from a single-view line drawing with sketch errors, which can handle a wide range of 3D man-made objects including curved components.
- 3
- DREAM: Decoupled Discriminative Learning with Bigraph-aware Alignment for Semi-supervised 2D-3D Cross-modal Retrieval3
The core of the DREAM is to decouple the label prediction and reliability measurement processes to reduce overconfident samples in discriminative learning and build a bigraph graph that connects cross-modal similar examples and then conduct learning to cluster with most edges kept for alignment.
- 3
- 3D Landmark Detection on Human Point Clouds: A Benchmark and A Dual Cascade Point Transformer Framework3
A Dual Cascade Point Transformer (D-CPT) model for precise point-based landmark detection is proposed, which gradually refines the landmarks through cascade Transformer decoder layers across the entire point cloud stream, simultaneously enhancing landmark coordinates with a RefineNet over local regions.
- 3
- Learning the Meta Feature Transformer for Unsupervised Person Re-Identification3
This work proposes a conceptually simple yet effective and learnable module effective block, named the meta feature transformer (MFT), which is a streamlined, lightweight network architecture that operates without the need for complex networks or feature memory bank storage.
- Self-ensembling for 3D point cloud domain adaptation3
An end-to-end self-ensembling network (SEN) for 3D point cloud domain adaption tasks is proposed, which outperforms the state-of-the-art methods on both classification and segmentation tasks and achieves better reconstruction results.
- 3
- Tucker Decomposition-Enhanced Dynamic Graph Convolutional Networks for Crowd Flows Prediction3
This study proposes a multi-scale fusion GCN-based framework with Tucker decomposition named mTDNet to enhance dynamic GCN for crowd flows prediction and utilizes the Tucker decomposition scheme to decompose ALT, allowing for effective capturing of the dynamic changes in crowd flow, including abnormal changes.
- A novel 25-ferroptosis-related gene signature for the prognosis of gliomas3
The ferroptosis-related genes play an important role in the malignant process of gliomas, potentially contributing to the development of prognostic stratification and treatment strategies.
- Cascaded Vehicle Matching and Short-Term Spatial-Temporal Network for Smoky Vehicle Detection3
A novel efficient cascaded framework for smoky vehicle detection which largely integrates prior knowledge and advanced deep networks is presented which is significantly superior to hand-crafted feature based methods and recent advanced methods.
- Video-based Smoky Vehicle Detection with A Coarse-to-Fine Framework3
A real-world large-scalesmoky vehicle dataset with 75,000 annotated smoky vehicle images is introduced, facilitating the effective training of advanced deep learning models and a new Coarse-to-fine Deep Smoky vehicle detection (CoDeS) framework is presented, significantly superior to those hand-crafted feature based methods and recent advanced methods.
- 3
- 3
- DPDEdit: Detail-Preserved Diffusion Models for Multimodal Fashion Image Editing2
This work presents Detail-Preserved Diffusion Models (DPDEdit), a new multimodal fashion image editing architecture based on latent diffusion models that outperforms state-of-the-art methods in terms of image fidelity and coherence with the given multimodal input.
- ResearchGPT: Benchmarking and Training LLMs for End-to-End Computer Science Research Workflows2
This work contributes CS-54k, a high-quality corpus of scientific Q&A pairs in computer science, built from 14k CC-licensed papers, and derives two complementary subsets: CS-4k, a carefully curated benchmark for evaluating AI's ability to assist scientific research, and CS-50k, a large-scale training dataset.
- 2
- MIMIC: Mask Image Pre-training with Mix Contrastive Fine-tuning for Facial Expression Recognition2
This work introduces a novel FER training paradigm named Mask Image pre-training with MIx Contrastive fine-tuning (MIMIC), and demonstrates that the MIMIC outperforms the previous training paradigm, showing its capability to learn better representations.
- 2
- ARFGCN: Adaptive Receptive Field Graph Convolutional Network for Urban Crowd Flow Prediction2
An Adaptive Receptive Field Graph Convolutional Network (ARFGCN) is proposed for urban crowd flow prediction that allows each region to independently determine its receptive field size, adaptively adjusted and learned in an end-to-end manner during training, enhancing model prediction performance.
- 2
- Phenotypic Screening of Molecular Docking Enriched Chemical Libraries from Targets Identified in Ischemic Stroke Genome Data by Network-Based Method2
This work aims to identify IS-specific targets by RNA sequencing data with a network-based approach and is expected to produce selective multidrug lead compounds for the development of treatments for complex diseases, such as ischemic stroke.
- 2
- 1
- 1
- 1
- TSFNet: A Temporal–Spectral Fusion Network for advanced speech emotion recognition in medical applications1
This study introduces a straightforward yet highly efficient network called TSFNet, which is the Temporal-Spectral Fusion Network via a Large-scale Pre-trained Model, specifically designed to effectively process intricate emotional nuances by seamlessly integrating temporal and spectral information present in speech signals.
- 1
- AU-Supervised Convolutional Vision Transformers for Synthetic Facial Expression Recognition1
A conceptually-new framework is proposed, termed as AU-Supervised Convolutional Vision Transformers (AU-CVT), which clearly improves the performance of FER by jointly training auxiliary datasets with AU or pseudo AU labels.
- 1
- 1
- Vision-and-Language Navigation: A Component-Centric Survey of Interactions, Coupling, and Deployment–
This survey revisits VLN from a component-internal perspective, viewing it as a navigation system composed of internal components such as instructions, environment representations, and embodied agents, and organizing existing work around the functions and interactions of these components.
- D2former: Dual-branch differential transformer for infrared image denoising–
D2former significantly enhances the quality of infrared images by adaptively applying differential mechanisms into spatial and channel attention computations and consistently outperforms state-of-the-art denoising methods in terms of both visual quality and structural fidelity.
- –
- –
- –
- –
- –
- Is Graph Convolution Always Beneficial For Every Feature?–
Compared to original GNNs, GFS significantly improves the extraction of useful topological information from each feature with comparable computational costs, and is robust to hyperparameter tuning, highlighting its potential as a universal method for enhancing various GNN architectures.
- –
- –
- –
- –
- –
- –
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 2026-10-12. 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.