Publications

You can also find my articles on my Google Scholar profile.

First-Author Papers

1. Twins: Revisiting the design of spatial attention in vision transformers, NeurIPS21
2. Conditional Positional Encodings for Vision Transformers, ICLR23
3. VisionLLaMA: A Unified LLaMA Backbone for Vision Tasks, ECCV24
4. GPG: A simple and strong reinforcement learning baseline for model reasoning, ICLR26
5. MobileVLM: A Fast, Strong and Open Vision Language Assistant for Mobile Devices
6. MobileVLM V2: Faster and Stronger Baseline for Vision Language Model
7. FairNAS: Rethinking evaluation fairness of weight sharing neural architecture search, ICCV21
8. Fair DARTS: Eliminating unfair advantages in differentiable architecture search, ECCV20
9. DARTS-: Robustly stepping out of performance collapse without indicators, ICLR21
10. ROME: Robustifying memory-efficient NAS via topology disentanglement and gradients accumulation, ICCV23
11. Make RepVGG Greater Again: A Quantization-aware Approach, AAAI24
12. MixPATH: A unified approach for one-shot neural architecture search, ICCV23
13. USP: Unified self-supervised pretraining for image generation and understanding, ICCV25
14. Noisy differentiable architecture search, BMVC21
15. A Unified Mixture-View Framework for Unsupervised Representation Learning, BMVC22
16. Multi-objective reinforced evolution in mobile neural architecture search, ECCVW2020
17. Fast, accurate and lightweight super-resolution with neural architecture search, ICPR20
18. MoGA: Searching beyond MobileNetV3, ICASSP2020
19. Scarlet-NAS: Bridging the gap between stability and scalability in weight-sharing NAS, ICCVW21
20. Parameter sharing deep deterministic policy gradient for cooperative multi-agent reinforcement learning
21. Improved crowding distance for NSGA-II
22. Policy optimization with penalized point probability distance: An alternative to PPO

Collaborative Papers

Image Generation & Editing

1. Layer-wise Instance Binding for Regional and Occlusion Control in Text-to-Image Diffusion Transformers, CVPR26
2. Geometry-Guided Reinforcement Learning for Multi-view Consistent 3D Scene Editing
3. From Scale to Speed: Adaptive Test-Time Scaling for Image Editing, CVPR26
4. Taming Preference Mode Collapse via Directional Decoupling Alignment in Diffusion Reinforcement Learning
5. From editor to dense geometry estimator, CVPR26
6. Ragsr: Regional attention guided diffusion for image super-resolution
7. S-Guidance: Stochastic Self Guidance for Training-Free Enhancement of Diffusion Models, ICLR26
8. LD-RPS: Zero-Shot Unified Image Restoration via Latent Diffusion Recurrent Posterior Sampling, ICCV25
9. Flux-text: A simple and advanced diffusion transformer baseline for scene text editing
10. Preference Alignment for Diffusion Model via Explicit Denoised Distribution Estimation
11. FlowDreamer: exploring high fidelity text-to-3D generation via rectified flow
12. TEXTS-Diff: TEXTS-Aware Diffusion Model for Real-World Text Image Super-Resolution, ICASSP26
13. Accurate and efficient single image super-resolution with matrix channel attention network, ACCV20

Video Generation & Understanding

1. Video-CoE: Reinforcing Video Event Prediction via Chain of Events, CVPR26
2. Taming Hallucinations: Boosting MLLMs' Video Understanding via Counterfactual Video Generation
3. Eevee: Towards Close-up High-resolution Video-based Virtual Try-on, CVPR26 Findings
4. ImagerySearch: Adaptive Test-Time Search for Video Generation Beyond Semantic Dependency Constraints, AAAI26
5. Video-star: Reinforcing open-vocabulary action recognition with tools, ICLR26
6. Omni-effects: Unified and spatially-controllable visual effects generation, AAAI26
7. Narrlv: Towards a comprehensive narrative-centric evaluation for long video generation models, ICLR26
8. VMBench: A Benchmark for Perception-Aligned Video Motion Generation, ICCV25
9. FingER: Content Aware Fine-grained Evaluation with Reasoning for AI-Generated Videos, ACM MM25
10. Latent Temporal Discrepancy as Motion Prior: A Loss-Weighting Strategy for Dynamic Fidelity in T2V, ICASSP26
11. Artifact-Aware Evaluation for High-Quality Video Generation, ICASSP26
12. Modeling Motion with Multi-Modal Features for Text-Based Video Segmentation, CVPR22

LLM Reasoning & Agents

1. Code2World: A GUI World Model via Renderable Code Generation
2. Entropy-Guided Data-Efficient Training for Multimodal Reasoning Reward Models
3. Harder Is Better: Boosting Mathematical Reasoning via Difficulty-Aware GRPO and Multi-Aspect Question Reformulation, ICLR26
4. AdaCuRL: Adaptive Curriculum Reinforcement Learning with Invalid Sample Mitigation and Historical Revisiting, AAAI26
5. Tree search for LLM agent reinforcement learning, ICLR26
6. AutoDrive-R: Incentivizing Reasoning and Self-Reflection Capacity for VLA Model in Autonomous Driving, ICLR26
7. Position bias mitigates position bias: Mitigate position bias through inter-position knowledge distillation, EMNLP25 oral
8. HS-STAR: Hierarchical Sampling for Self-Taught Reasoners via Difficulty Estimation and Budget Reallocation, EMNLP25 oral
9. Ranking-aware Reinforcement Learning for Ordinal Ranking, ICASSP26

Multimodal & Vision-Language

1. What if Agents Could Imagine? Reinforcing Open-Vocabulary HOI Comprehension through Generation
2. Q-Hawkeye: Reliable Visual Policy Optimization for Image Quality Assessment
3. Everything in Its Place: Benchmarking Spatial Intelligence of Text-to-Image Models, ICLR26
4. Urban Socio-Semantic Segmentation with Vision-Language Reasoning, ICLR26
5. Where and What Matters: Sensitivity-Aware Task Vectors for Many-Shot Multimodal In-Context Learning, AAAI26
6. Univg-r1: Reasoning guided universal visual grounding with reinforcement learning
7. Next Token Is Enough: Realistic Image Quality and Aesthetic Scoring with Multimodal Large Language Model
8. Mmgenbench: Evaluating the limits of LMMs from the text-to-image generation perspective
9. Lenna: Language Enhanced Reasoning Detection Assistant, ICASSP25

Detection, Segmentation & 3D Perception

1. UPRE: Zero-Shot Domain Adaptation for Object Detection via Unified Prompt and Representation Enhancement, ICCV25
2. PLUG: Revisiting Amodal Segmentation with Foundation Model and Hierarchical Focus, CVPR25
3. SCTNet: Single Branch CNN with Transformer Semantic Information for Real-time Segmentation, AAAI24
4. FastPillars: A Deployment-friendly Pillar-based 3D Detector, IEEE TCSVT
5. Yolov6 v3.0: A full-scale reloading
6. AeDet: Azimuth-invariant multi-view 3D object detection, CVPR23
7. SegViT: Semantic segmentation with plain vision transformers, NeurIPS22
8. YOLOv6: A single-stage object detection framework for industrial applications, arXiv
9. Fully convolutional one-stage 3D object detection on LiDAR range images, NeurIPS22
10. PromptDet: Towards open-vocabulary detection using uncurated images, ECCV22
11. Cctrans: Simplifying and improving crowd counting with transformer

Foundation Model Architectures

1. FASA: Frequency-Aware Sparse Attention, ICLR26
2. AR-MAP: Are Autoregressive Large Language Models Implicit Teachers for Diffusion Large Language Models?
3. Semantic Context Matters: Improving Conditioning for Autoregressive Models
4. There is No VAE: End-to-End Pixel-Space Generative Modeling via Self-Supervised Pre-training, ICLR26
5. Scalar: Scale-wise controllable visual autoregressive learning, AAAI26
6. Dyn-Adapter: Towards Disentangled Representation for Efficient Visual Recognition, ECCV24
7. Revealing the Dark Secrets of Extremely Large Kernel ConvNets on Robustness, ICML24
8. PeLK: Parameter-efficient Large Kernel ConvNets with Peripheral Convolution, CVPR24
9. Efficientrep: an efficient repvgg-style convnets with hardware-aware neural network design

Model Compression & AutoML

1. LiDAR-PTQ: Post-Training Quantization for Point Cloud 3D Object Detection, ICLR24
2. Masked Autoencoders Are Robust Neural Architecture Search Learners
3. A Speed Odyssey for Deployable Quantization of LLMs
4. Norm Tweaking: High-performance Low-bit Quantization of Large Language Models, AAAI24
5. FPTQ: Fine-grained Post-Training Quantization for Large Language Models
6. EAPruning: Evolutionary Pruning for Vision Transformers and CNNs, BMVC22
7. DAAS: Differentiable architecture and augmentation policy search
8. AutoKWS: Keyword Spotting with Differentiable Architecture Search, ICASSP21
9. Neural Architecture Search on Acoustic Scene Classification, InterSpeech20

Maps, Mobility & Recommendation

1. MobilityBench: A Scalable Benchmark for Evaluating Route-Planning Agents in Real-World Mobility Scenarios
2. IntRR: A Framework for Integrating SID Redistribution and Length Reduction for Generative Recommendation
3. IntTravel: A Real-World Dataset and Generative Framework for Integrated Multi-Task Travel Recommendation
4. GenMRP: A Generative Multi-Route Planning Framework for Efficient and Personalized Real-Time Industrial Navigation
5. SCASRec: A Self-Correcting and Auto-Stopping Model for Generative Route List Recommendation
6. Thinking with Map: Reinforced Parallel Map-Augmented Agent for Geolocalization
7. Intsr: An integrated generative framework for search and recommendation
8. Comprehensive Comparison Network: a framework for locality-aware, routes-comparable and interpretable route recommendation
9. Effective Probabilistic Time Series Forecasting with Fourier Adaptive Noise-Separated Diffusion
10. DSFNet: Learning Disentangled Scenario Factorization for Multi-Scenario Route Ranking, WWW25