Jiaying Liu
教授
Leads the STRUCT group, researching image and video restoration, compression and visual understanding.
Jiaying Liu's lab at Peking works on recognition, detection & segmentation, video understanding and efficient training & inference. 69 papers since 2023, 23 at top venues, mostly CVPR and AAAI. 11 PhD students. 18 alumni, most often to Tencent and Apple. Applications close .
At a glance
- papers, 2023–2026
- 18/ year
- at T1 venues
- 33%
- PhD students
- 11
- MS students
- 0
- alumni listed
- 18
Papers by yearfrom the lab homepage, 2023–2026
- 19 papers2026 · T1 4 · collecting
- 12 papers2025 · T1 7
- 22 papers2024 · T1 9
- 16 papers2023 · T1 3
T1 venuesOtherVenues: CVPR, AAAI, ACM MM, ECCV, ICCV
Research focuspaper titles that use each topic's words
- Video understanding16 papers
- Recognition, detection & segmentation9 papers
- Efficient training & inference3 papers
- video restoration0 papers
- visual understanding0 papers
Paper list
From the lab homepage.
202619 papers
- Accepted by ACM MultimediaBridging the Reference Gap: Dual-Reference SR-IQA with Generated Image Guidance as Perceptual Anchors
- Accepted by Journal of SoftwareCausal Inference-Driven Image Compression via World Models
- Proc. of IEEE International Conference on Multimedia & ExpoConnecting Generation with Compression: Unified Generative Coding with Semantic Guidance
- Accepted by IEEE Trans. on MultimediaDiffSSR+: Optimized Progressive Diffusion with Stereo-Consistent Sampling for Stereo Super-Resolution
- IEEE Trans. on MultimediaDiffTile: High-Resolution Periodic Pattern Generation via Splice, Focus and Enhance
- Proc. of IEEE International Conference on Acoustics, Speech, and Signal ProcessingEnhanced Video Compression with Context-Aware Dynamic Neural Adapter
- CVPR (T1)FlowPortal: Residual-Corrected Flow for Training-Free Video Relighting and Background Replacement
- Accepted by Advances in Neural Information Processing SystemsIris: Empowering Video MLLMs with High-Frequency Pose Priors via Spatiotemporal Binding
- IEEE Trans. on Image ProcessingLocal Dimension Enhancement Representation Learning for Skeleton-Based Action Segmentation
- IEEE Trans. on Image ProcessingPrompting Rain Off: Evolving Compact Dual Prompts for Continual De-Raining
- IEEE Trans. on Pattern Analysis and Machine IntelligenceQuadPrior++: Multi-dimension Augmented Physical Prior for Zero-Reference Illumination Enhancement
- Accepted by ACM Trans. on Multimedia Computing Communications and ApplicationsSeeing in the Dark with Ambient Guidance
202512 papers
- Journal of Computer-Aided Design & Computer GraphicsAIGC-Based Image and Video Generation Method: A Review
- ICCV (T1)AnyPortal: Zero-Shot Consistent Video Background Replacement
- ICCV (T1)Cross-Granularity Online Optimization with Masked Compensated Information for Learned Image Compression
- IEEE Trans. on Circuits and Systems for Video TechnologyFacial Image Compression via Neural Image Manifold Compression
- CVPR (T1)JanusFlow: Harmonizing Autoregression and Rectified Flow for Unified Multimodal Understanding and Generation
- Proc. of IEEE International Conference on Multimedia & ExpoProgressive Interactive Diffusion Model for Real-World Stereo Super-Resolution
- CVPR (T1)PTDiffusion: Free Lunch for Generating Optical Illusion Hidden Pictures using Phase-Transferred Diffusion Model
- IEEE Trans. on Pattern Analysis and Machine IntelligenceSelf-Supervised Skeleton Representation Learning via Actionlet Contrast and Reconstruct
- NeurIPS (T1)SGAR: Structural Generative Augmentation for 3D Human Motion Retrieval
- Proc. of IEEE International Symposium on Circuits and SystemsSplice, Focus and Relife: High-Resolution Periodic Pattern Generation
- IJCV (T1)Swap Attention in Spatiotemporal Diffusions for Text-to-Video Generation
- AAAI (T1)UP-Restorer: When Unrolling Meets Prompts for Unified Image Restoration
202422 papers
- ACM MMCOCO-LC: Colorfulness Controllable Language-based Colorization
- ACM MMConsistency Guided Diffusion Model with Neural Syntax for Perceptual Image Compression
- Proc. of ACM Multimedia AsiaCoolColor: Text-guided COherent OLd film COLORization
- ICML (T1)Correcting Diffusion-Based Perceptual Image Compression with Privileged End-to-End Decoder
- IEEE Trans. on Geoscience and Remote SensingDiffusion Enhancement for Cloud Removal in Ultra-Resolution Remote Sensing Imagery
- ACM MMFBSDiff: Plug-and-Play Frequency Band Substitution of Diffusion Features for Highly Controllable Text-Driven Image Translation
- AAAI (T1)Frequency-Controlled Diffusion Model for Versatile Text-Guided Image-to-Image Translation
- CVPR (T1)FRESCO: Spatial-Temporal Correspondence for Zero-Shot Video Translation
- ECCV (T1)Idempotent Unsupervised Representation Learning for Skeleton-Based Action Recognition
- APSIPA Trans. on Signal and Information ProcessingIntelligent Artistic Typography: A Comprehensive Review of Artistic Text Design and Generation
- IEEE Trans. on Pattern Analysis and Machine IntelligenceLearning to Remove Rain in Video with Self-Supervision
- ECCV (T1)MacDiff: Unified Skeleton Modeling with Masked Conditional Diffusion
202316 papers
- CVPR (T1)Actionlet-Dependent Contrastive Learning for Unsupervised Skeleton-Based Action Recognition
- IEEE Trans. on Industrial InformaticsAutomatic Model-Based Dataset Generation for High-Level Vision Tasks of Autonomous Driving in Haze Weather
- Proc. of IEEE International Symposium on Circuits and SystemsBayesian Contrastive Learning with Manifold Regularization for Self-Supervised Skeleton Based Action Recognition
- Proc. of IEEE International Conference on Multimedia & ExpoCollaborative Spatial-Temporal Distillation for Efficient Video Deraining
- ICIPContent-Adaptive Parallel Entropy Coding for End-to-End Image Compression
- ACM Trans. on Multimedia Computing Communications and ApplicationsDeep Inter Prediction with Error-Corrected Auto-Regressive Network for Video Coding
- Proc. of International Joint Conferences on Artificial IntelligenceDual Prompt Learning for Continual Rain Removal from Single Images
- APSIPA Trans. on Signal and Information ProcessingEMS^2L: Enhanced Multi-Task Self-Supervised Learning for 3D Skeleton Representation Learning
- ICIPFlash Compensated Low-Light Enhancement via Hierarchical Network Prediction
- AAAI (T1)Hierarchical Consistent Contrastive Learning for Skeleton-Based Action Recognition with Growing Augmentations
- IEEE Trans. on MultimediaIntelligent Typography: Artistic Text Style Transfer for Complex Texture and Structure
- ACM MMPrompted Contrast with Masked Motion Modeling: Towards Versatile 3D Action Representation Learning
Where alumni went18 listed, 69% to industry
Industry 9Academia 4
Most often to
- Tencent
- Apple
- Alibaba
Also to
- City University of Hong Kong
- Institute of Automation Chinese Academy of Sciences
- NTU Singapore
Apply to PekingFall 2027
Computer Science PhD (international applicants)
- Round
- Main
- Opens
- —
- GRE
- See program page
- English
- See program page
- Fee
- See program page
Not yet confirmed for this intake.
Similar labsnearest by research summary
- USUIUC
Pierre Moulin
- Recognition, detection & segmentation
- CNSJTU
Tanfeng Sun
- Recognition, detection & segmentation
- CNPeking
Liangcai Gao
- Recognition, detection & segmentation
- SGNUS
Tsuhan Chen
- Recognition, detection & segmentation
- SGNTU
Bihan Wen
- Computational imaging & photography
- CNSJTU
Weixia Zhang
- Computational imaging & photography
Your notes
Save this lab to keep notes.
Data
Checked against the lab homepage.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Lab homepage
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.
Pekingcloses