Yunfan Liu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Yunfan Liu has a total of 46 co-authors. Here are the top co-authors by number of shared publications:
- Zhenan Sun
Institute of Automation, Chinese Academy of Sciences, Beijing, China
15 shared publications - Qi Li
School of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing, China
10 shared publications Qi Li
Center for Research on Intelligent Perception and Computing CASIA
6 shared publications- Qixiang Ye
University of Chinese Academy of Sciences
5 shared publications - Qiyao Deng
Peoples Public Security University of China, Beijing, China
4 shared publications - Jie Cao
Institute of Automation, Chinese Academy of Sciences
3 shared publications
Latest Publications
- 3 September 2026
- Published by Springer Nature in Machine Intelligence Research
- Vol. 23  (5) , 1216-1228
StyleVideo: One-shot Image-guided Video Style Transfer
- 3 September 2026
- Published by Springer Nature in Machine Intelligence Research
- Vol. 23  (5) , 1216-1228
StyleVideo: One-shot Image-guided Video Style Transfer
- Open Access
Balancing Understanding and Generation in Discrete Diffusion Models
In discrete generative modeling, two dominant paradigms demonstrate divergent capabilities: Masked Diffusion Language Models (MDLM) excel at semantic understanding and zero-shot generalization, whereas Uniform-noise Diffusion Language Models (UDLM) achieve strong few-step generation quality, yet neither attains balanced performance across both dimensions. To address this, we propose XDLM, which bridges the two paradigms via a stationary noise kernel. XDLM offers two key contributions: (1) it provides a principled theoretical unification of MDLM and UDLM, recovering each paradigm as a special case; and (2) an alleviated memory bottleneck enabled by an algebraic simplification of the posterior probabilities. Experiments demonstrate that XDLM advances the Pareto frontier between understanding capability and generation quality. Quantitatively, XDLM surpasses UDLM by 5.4 points on zero-shot text benchmarks and outperforms MDLM in few-step image generation (FID 54.1 vs. 80.8). When scaled to tune an 8B-parameter large language model, XDLM achieves 15.0 MBPP in just 32 steps, effectively doubling the baseline performance. Finally, analysis of training dynamics reveals XDLM's superior potential for long-term scaling. Code is available at https://github.com/MzeroMiko/XDLM
Open AccessBalancing Understanding and Generation in Discrete Diffusion Models
In discrete generative modeling, two dominant paradigms demonstrate divergent capabilities: Masked Diffusion Language Models (MDLM) excel at semantic understanding and zero-shot generalization, whereas Uniform-noise Diffusion Language Models (UDLM) achieve strong few-step generation quality, yet neither attains balanced performance across both dimensions. To address this, we propose XDLM, which bridges the two paradigms via a stationary noise kernel. XDLM offers two key contributions: (1) it provides a principled theoretical unification of MDLM and UDLM, recovering each paradigm as a special case; and (2) an alleviated memory bottleneck enabled by an algebraic simplification of the posterior probabilities. Experiments demonstrate that XDLM advances the Pareto frontier between understanding capability and generation quality. Quantitatively, XDLM surpasses UDLM by 5.4 points on zero-shot text benchmarks and outperforms MDLM in few-step image generation (FID 54.1 vs. 80.8). When scaled to tune an 8B-parameter large language model, XDLM achieves 15.0 MBPP in just 32 steps, effectively doubling the baseline performance. Finally, analysis of training dynamics reveals XDLM's superior potential for long-term scaling. Code is available at https://github.com/MzeroMiko/XDLM
- 12 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biometrics, Behavior, and Identity Science
- Vol. 8  (2) , 244-254
Affinity Contrastive Learning for Skeleton-Based Human Activity Understanding
In skeleton-based human activity understanding, existing methods often adopt the contrastive learning paradigm to construct a discriminative feature space. However, many of these approaches fail to exploit the structural inter-class similarities and overlook the impact of anomalous positive samples. In this study, we introduce ACLNet, an Affinity Contrastive Learning Network that explores the intricate clustering relationships among human activity classes to improve feature discrimination. Specifically, we propose an affinity metric to refine similarity measurements, thereby forming activity superclasses that provide more informative contrastive signals. A dynamic temperature schedule is also introduced to adaptively adjust the penalty strength for various superclasses. In addition, we employ a margin-based contrastive strategy to improve the separation of hard positive and negative samples within classes. Extensive experiments on NTU RGB+D 60, NTU RGB+D 120, Kinetics-Skeleton, PKU-MMD, FineGYM, and CASIA-B demonstrate the superiority of our method in skeleton-based action recognition, gait recognition, and person re-identification. The source code is available at https://github.com/firework8/ACLNet
- 12 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biometrics, Behavior, and Identity Science
- Vol. 8  (2) , 244-254
Affinity Contrastive Learning for Skeleton-Based Human Activity Understanding
In skeleton-based human activity understanding, existing methods often adopt the contrastive learning paradigm to construct a discriminative feature space. However, many of these approaches fail to exploit the structural inter-class similarities and overlook the impact of anomalous positive samples. In this study, we introduce ACLNet, an Affinity Contrastive Learning Network that explores the intricate clustering relationships among human activity classes to improve feature discrimination. Specifically, we propose an affinity metric to refine similarity measurements, thereby forming activity superclasses that provide more informative contrastive signals. A dynamic temperature schedule is also introduced to adaptively adjust the penalty strength for various superclasses. In addition, we employ a margin-based contrastive strategy to improve the separation of hard positive and negative samples within classes. Extensive experiments on NTU RGB+D 60, NTU RGB+D 120, Kinetics-Skeleton, PKU-MMD, FineGYM, and CASIA-B demonstrate the superiority of our method in skeleton-based action recognition, gait recognition, and person re-identification. The source code is available at https://github.com/firework8/ACLNet
- 6 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 63  (01962892) , 1-16
CC-Diff++: Spatially Controllable Text-to-Image Synthesis for Remote Sensing With Enhanced Contextual Coherence
Generating visually realistic remote sensing (RS) images requires maintaining semantic coherence between objects and their surrounding environments. However, existing image synthesis methods prioritize foreground controllability while oversimplifying backgrounds into plain or generic textures. This oversight neglects the crucial interaction between foreground and background elements, resulting in semantic inconsistencies in RS scenarios. To address this challenge, we propose CC-Diff++, a Diffusion Model-based approach for spatially controllable RS image synthesis with enhanced Context Coherence. To capture spatial interdependence, we propose a novel module named Co-Resampler, which employs an advanced masked attention mechanism to jointly extract features from both the foreground and background while modeling their mutual relationships. Furthermore, we introduce a text-to-layout prediction module powered by Large Language Models (LLMs) and a reference image retrieval mechanism for providing rich textural guidance, which work together to enable CC-Diff++ to generate outputs that are both more diverse and more realistic. Extensive experiments demonstrate that CC-Diff++ outperforms state-of-the-art methods in visual fidelity, semantic accuracy, and positional precision on multiple RS datasets. CC-Diff++ also shows strong trainability, improving detection accuracy by 2.04 mAP on DOTA and 11.81 mAP on the HRSC dataset.
- 6 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 63  (01962892) , 1-16
CC-Diff++: Spatially Controllable Text-to-Image Synthesis for Remote Sensing With Enhanced Contextual Coherence
Generating visually realistic remote sensing (RS) images requires maintaining semantic coherence between objects and their surrounding environments. However, existing image synthesis methods prioritize foreground controllability while oversimplifying backgrounds into plain or generic textures. This oversight neglects the crucial interaction between foreground and background elements, resulting in semantic inconsistencies in RS scenarios. To address this challenge, we propose CC-Diff++, a Diffusion Model-based approach for spatially controllable RS image synthesis with enhanced Context Coherence. To capture spatial interdependence, we propose a novel module named Co-Resampler, which employs an advanced masked attention mechanism to jointly extract features from both the foreground and background while modeling their mutual relationships. Furthermore, we introduce a text-to-layout prediction module powered by Large Language Models (LLMs) and a reference image retrieval mechanism for providing rich textural guidance, which work together to enable CC-Diff++ to generate outputs that are both more diverse and more realistic. Extensive experiments demonstrate that CC-Diff++ outperforms state-of-the-art methods in visual fidelity, semantic accuracy, and positional precision on multiple RS datasets. CC-Diff++ also shows strong trainability, improving detection accuracy by 2.04 mAP on DOTA and 11.81 mAP on the HRSC dataset.
- 8 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. 37  (1) , 191-203
Expandable Residual Approximation for Knowledge Distillation
Knowledge distillation (KD) aims to transfer knowledge from a large-scale teacher model to a lightweight one, significantly reducing computational and storage requirements. However, the inherent learning capacity gap between the teacher and student often hinders the sufficient transfer of knowledge, motivating numerous studies to address this challenge. Inspired by the progressive approximation principle in the Stone–Weierstrass theorem, we propose expandable residual approximation (ERA), a novel KD method that decomposes the approximation of residual knowledge into multiple steps, reducing the difficulty of mimicking the teacher’s representation through a divide-and-conquer approach. Specifically, ERA employs a multibranched residual network (MBRNet) to implement this residual knowledge decomposition. Additionally, a teacher weight integration (TWI) strategy is introduced to mitigate the capacity disparity by reusing the teacher’s head weights. Extensive experiments show that ERA improves the Top-1 accuracy on ImageNet classification benchmark by 1.41% and the AP on the MS COCO object detection benchmark by 1.40, as well as achieving leading performance across computer vision tasks.
- 8 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. 37  (1) , 191-203
Expandable Residual Approximation for Knowledge Distillation
Knowledge distillation (KD) aims to transfer knowledge from a large-scale teacher model to a lightweight one, significantly reducing computational and storage requirements. However, the inherent learning capacity gap between the teacher and student often hinders the sufficient transfer of knowledge, motivating numerous studies to address this challenge. Inspired by the progressive approximation principle in the Stone–Weierstrass theorem, we propose expandable residual approximation (ERA), a novel KD method that decomposes the approximation of residual knowledge into multiple steps, reducing the difficulty of mimicking the teacher’s representation through a divide-and-conquer approach. Specifically, ERA employs a multibranched residual network (MBRNet) to implement this residual knowledge decomposition. Additionally, a teacher weight integration (TWI) strategy is introduced to mitigate the capacity disparity by reusing the teacher’s head weights. Extensive experiments show that ERA improves the Top-1 accuracy on ImageNet classification benchmark by 1.41% and the AP on the MS COCO object detection benchmark by 1.40, as well as achieving leading performance across computer vision tasks.
- Open Access
Expandable Residual Approximation for Knowledge Distillation
Knowledge distillation (KD) aims to transfer knowledge from a large-scale teacher model to a lightweight one, significantly reducing computational and storage requirements. However, the inherent learning capacity gap between the teacher and student often hinders the sufficient transfer of knowledge, motivating numerous studies to address this challenge. Inspired by the progressive approximation principle in the Stone-Weierstrass theorem, we propose Expandable Residual Approximation (ERA), a novel KD method that decomposes the approximation of residual knowledge into multiple steps, reducing the difficulty of mimicking the teacher's representation through a divide-and-conquer approach. Specifically, ERA employs a Multi-Branched Residual Network (MBRNet) to implement this residual knowledge decomposition. Additionally, a Teacher Weight Integration (TWI) strategy is introduced to mitigate the capacity disparity by reusing the teacher's head weights. Extensive experiments show that ERA improves the Top-1 accuracy on the ImageNet classification benchmark by 1.41% and the AP on the MS COCO object detection benchmark by 1.40, as well as achieving leading performance across computer vision tasks. Codes and models are available at https://github.com/Zhaoyi-Yan/ERA.
Open AccessExpandable Residual Approximation for Knowledge Distillation
Knowledge distillation (KD) aims to transfer knowledge from a large-scale teacher model to a lightweight one, significantly reducing computational and storage requirements. However, the inherent learning capacity gap between the teacher and student often hinders the sufficient transfer of knowledge, motivating numerous studies to address this challenge. Inspired by the progressive approximation principle in the Stone-Weierstrass theorem, we propose Expandable Residual Approximation (ERA), a novel KD method that decomposes the approximation of residual knowledge into multiple steps, reducing the difficulty of mimicking the teacher's representation through a divide-and-conquer approach. Specifically, ERA employs a Multi-Branched Residual Network (MBRNet) to implement this residual knowledge decomposition. Additionally, a Teacher Weight Integration (TWI) strategy is introduced to mitigate the capacity disparity by reusing the teacher's head weights. Extensive experiments show that ERA improves the Top-1 accuracy on the ImageNet classification benchmark by 1.41% and the AP on the MS COCO object detection benchmark by 1.40, as well as achieving leading performance across computer vision tasks. Codes and models are available at https://github.com/Zhaoyi-Yan/ERA.
- 30 May 2024
- Published by Springer Nature in Machine Intelligence Research
- Vol. 21  (5) , 941-953
One-shot Face Reenactment with Dense Correspondence Estimation
- 30 May 2024
- Published by Springer Nature in Machine Intelligence Research
- Vol. 21  (5) , 941-953
One-shot Face Reenactment with Dense Correspondence Estimation
- Open Access
- 24 March 2024
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 38  (7) , 6782-6790
Spatial Transform Decoupling for Oriented Object Detection
Vision Transformers (ViTs) have achieved remarkable success in computer vision tasks. However, their potential in rotation-sensitive scenarios has not been fully explored, and this limitation may be inherently attributed to the lack of spatial invariance in the data-forwarding process. In this study, we present a novel approach, termed Spatial Transform Decoupling (STD), providing a simple-yet-effective solution for oriented object detection with ViTs. Built upon stacked ViT blocks, STD utilizes separate network branches to predict the position, size, and angle of bounding boxes, effectively harnessing the spatial transform potential of ViTs in a divide-and-conquer fashion. Moreover, by aggregating cascaded activation masks (CAMs) computed upon the regressed parameters, STD gradually enhances features within regions of interest (RoIs), which complements the self-attention mechanism. Without bells and whistles, STD achieves state-of-the-art performance on the benchmark datasets including DOTA-v1.0 (82.24% mAP) and HRSC2016 (98.55% mAP), which demonstrates the effectiveness of the proposed method. Source code is available at https://github.com/yuhongtian17/Spatial-Transform-Decoupling.
- 24 March 2024
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 38  (7) , 6782-6790
Open AccessSpatial Transform Decoupling for Oriented Object Detection
Vision Transformers (ViTs) have achieved remarkable success in computer vision tasks. However, their potential in rotation-sensitive scenarios has not been fully explored, and this limitation may be inherently attributed to the lack of spatial invariance in the data-forwarding process. In this study, we present a novel approach, termed Spatial Transform Decoupling (STD), providing a simple-yet-effective solution for oriented object detection with ViTs. Built upon stacked ViT blocks, STD utilizes separate network branches to predict the position, size, and angle of bounding boxes, effectively harnessing the spatial transform potential of ViTs in a divide-and-conquer fashion. Moreover, by aggregating cascaded activation masks (CAMs) computed upon the regressed parameters, STD gradually enhances features within regions of interest (RoIs), which complements the self-attention mechanism. Without bells and whistles, STD achieves state-of-the-art performance on the benchmark datasets including DOTA-v1.0 (82.24% mAP) and HRSC2016 (98.55% mAP), which demonstrates the effectiveness of the proposed method. Source code is available at https://github.com/yuhongtian17/Spatial-Transform-Decoupling.
- Open Access
Spatial Transform Decoupling for Oriented Object Detection
Vision Transformers (ViTs) have achieved remarkable success in computer vision tasks. However, their potential in rotation-sensitive scenarios has not been fully explored, and this limitation may be inherently attributed to the lack of spatial invariance in the data-forwarding process. In this study, we present a novel approach, termed Spatial Transform Decoupling (STD), providing a simple-yet-effective solution for oriented object detection with ViTs. Built upon stacked ViT blocks, STD utilizes separate network branches to predict the position, size, and angle of bounding boxes, effectively harnessing the spatial transform potential of ViTs in a divide-and-conquer fashion. Moreover, by aggregating cascaded activation masks (CAMs) computed upon the regressed parameters, STD gradually enhances features within regions of interest (RoIs), which complements the self-attention mechanism. Without bells and whistles, STD achieves state-of-the-art performance on the benchmark datasets including DOTA-v1.0 (82.24% mAP) and HRSC2016 (98.55% mAP), which demonstrates the effectiveness of the proposed method. Source code is available at https://github.com/yuhongtian17/Spatial-Transform-Decoupling.
Open AccessSpatial Transform Decoupling for Oriented Object Detection
Vision Transformers (ViTs) have achieved remarkable success in computer vision tasks. However, their potential in rotation-sensitive scenarios has not been fully explored, and this limitation may be inherently attributed to the lack of spatial invariance in the data-forwarding process. In this study, we present a novel approach, termed Spatial Transform Decoupling (STD), providing a simple-yet-effective solution for oriented object detection with ViTs. Built upon stacked ViT blocks, STD utilizes separate network branches to predict the position, size, and angle of bounding boxes, effectively harnessing the spatial transform potential of ViTs in a divide-and-conquer fashion. Moreover, by aggregating cascaded activation masks (CAMs) computed upon the regressed parameters, STD gradually enhances features within regions of interest (RoIs), which complements the self-attention mechanism. Without bells and whistles, STD achieves state-of-the-art performance on the benchmark datasets including DOTA-v1.0 (82.24% mAP) and HRSC2016 (98.55% mAP), which demonstrates the effectiveness of the proposed method. Source code is available at https://github.com/yuhongtian17/Spatial-Transform-Decoupling.
- 19 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 19  (15566013) , 3093-3108
Cross-Scenario Unknown-Aware Face Anti-Spoofing With Evidential Semantic Consistency Learning
In recent years, domain adaptation techniques have been widely used to adapt face anti-spoofing models to a cross-scenario target domain. Most previous methods assume that the Presentation Attack Instruments (PAIs) in such cross-scenario target domain are same as in the source domain. However, as the malicious users are free to use any form of unknown PAIs to attack the system, this assumption does not always hold in practical applications of face anti-spoofing. Thus, unknown PAIs would inevitably lead to significant performance degradation, since samples of known and unknown PAIs usually have large differences. In this paper, we propose an Evidential Semantic Consistency Learning (ESCL) framework to address this problem. Specifically, a regularized evidential deep learning strategy with a two-way balance of class probability and uncertainty is leveraged to produce uncertainty scores for unknown PAI detection. Meanwhile, entropy optimization-based semantic consistency learning strategy is also employed to encourage features of live and known PAIs to be gathered in the label-conditioned clusters across the source and target domains, while make the features of unknown PAIs to be self-clustered according to intrinsic semantic information. In addition, a new evaluation metric, KUHAR, is proposed to comprehensively evaluate the error rate of known classes and unknown PAIs. Extensive experimental results on six public datasets demonstrate the effectiveness of our method in generalizing face anti-spoofing models to both known classes and unknown PAIs with different types and quantities in a cross-scenario testing domain. Our method achieves state-of-the-art performance on eight different protocols.
- 19 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 19  (15566013) , 3093-3108
Cross-Scenario Unknown-Aware Face Anti-Spoofing With Evidential Semantic Consistency Learning
In recent years, domain adaptation techniques have been widely used to adapt face anti-spoofing models to a cross-scenario target domain. Most previous methods assume that the Presentation Attack Instruments (PAIs) in such cross-scenario target domain are same as in the source domain. However, as the malicious users are free to use any form of unknown PAIs to attack the system, this assumption does not always hold in practical applications of face anti-spoofing. Thus, unknown PAIs would inevitably lead to significant performance degradation, since samples of known and unknown PAIs usually have large differences. In this paper, we propose an Evidential Semantic Consistency Learning (ESCL) framework to address this problem. Specifically, a regularized evidential deep learning strategy with a two-way balance of class probability and uncertainty is leveraged to produce uncertainty scores for unknown PAI detection. Meanwhile, entropy optimization-based semantic consistency learning strategy is also employed to encourage features of live and known PAIs to be gathered in the label-conditioned clusters across the source and target domains, while make the features of unknown PAIs to be self-clustered according to intrinsic semantic information. In addition, a new evaluation metric, KUHAR, is proposed to comprehensively evaluate the error rate of known classes and unknown PAIs. Extensive experimental results on six public datasets demonstrate the effectiveness of our method in generalizing face anti-spoofing models to both known classes and unknown PAIs with different types and quantities in a cross-scenario testing domain. Our method achieves state-of-the-art performance on eight different protocols.