Email    /    Scholar   /   Github   /   X

Houwen Peng (彭厚文)

Houwen is currently an independent researcher and developer specializing in multimodal generative AI systems.
Previously, he led Tencent’s Hunyuan multimodal teams, where models built under his oversight secured top-tier performance on LMArena leaderboard and serve hundreds of millions of daily user queries across Tencent's int/external applications, e.g., Yuanbao. Prior to Tencent, he was a Senior Researcher at Microsoft Research Asia (MSRA, 2018-2024). His research on AI foundations has been shipped as core technologies to various Microsoft key products, including Azure, Office, Bing, Visual Studio, etc. He has been recognized as one of world's top 2% scientist.

Email: penghouwen [at] icloud [dot] com
Wechat: [71, 27131, 13312, 40657]   (tokenized thru gpt-5.x)

News

Publications

R-4B: Incentivizing General-Purpose Auto-Thinking Capability in MLLMs via Bi-Mode Annealing and Reinforce Learning
2025 Computer Vision, Foundation Models Selected

Qi Yang, Bolin Ni, Shiming Xiang, Han Hu, Houwen Peng, Jie Jiang
Tech Report
Tech Report  •   Code  •   HF

R-BenchV: A Primary Assessment for Visual Reasoning Models with Multi-modal Outputs
2015 Multimodal Learning, Neural Architecture Search, Transformer Architecture

Meng-Hao Guo, Xuanyu Chu, Qianrui Yang, Zhe-Han Mo, Yiqing Shen, Pei-lin Li, Xinjie Lin, Jinnian Zhang, Xin-Sheng Chen, Yi Zhang, Kiyohiro Nakayama, Zhengyang Geng, Houwen Peng, Han Hu, Shi-min Hu
Tech Report
Tech Report  •   Code  •   Leaderboard

R-Bench: Graduate-level Multi-disciplinary Benchmarks for LLM & MLLM Complex Reasoning Evaluation
2020 Transformer Architecture, Natural Language Processing, Computer Vision, Large Language Models Selected

Meng-Hao Guo, Jiajun Xu, Yi Zhang, Jiaxi Song, Haoyang Peng, Yi-Xuan Deng, Xinzhi Dong, Kiyohiro Nakayama, Zhengyang Geng, Chen Wang, Bolin Ni, Yongming Rao, Houwen Peng, Han Hu, Gordon Wetzstein, Shi-min Hu
ICML 2025
Paper  •   Code  •   Leaderboard

Mitigating Visual Forgetting via Take-along Visual Conditioning for Multi-modal Long CoT Reasoning
2023 Transformer Architecture, Computer Vision, Foundation Models Selected

Hai-Long Sun, Zhun Sun, Houwen Peng, Han-Jia Ye
ACL 2025
Paper  •   Code

ScalingFilter: Assessing Data Quality through Inverse Utilization of Scaling Laws
2025 Neural Architecture Search, Multimodal Learning

Ruihang Li, Yixuan Wei, Miaosen Zhang, Nenghai Yu, Han Hu, Houwen Peng
EMNLP 2024
Paper  •   Code

Xwin-LM: Strong and Scalable Alignment Practice for LLMs
2015 Model Compression, Visual Tracking

Bolin Ni, Jingcheng Hu, Yixuan Wei, Houwen Peng, Zheng Zhang, Gaofeng Meng, Han Hu
Tech Report
Tech Report  •   Code

Common 7B Language Models Already Possess Strong Math Capabilities
2019 Multimodal Learning, Transformer Architecture, Natural Language Processing

Chen Li, Weiqi Wang, Jingcheng Hu, Yixuan Wei, Nanning Zheng, Han Hu, Zheng Zhang, Houwen Peng
Tech Report
Tech Report  •   Code

FP8-LM: Training FP8 Large Language Models
2022 Computer Vision, Natural Language Processing

Houwen Peng, Kan Wu, Yixuan Wei, Guoshuai Zhao, Yuxiang Yang, Ze Liu, Yifan Xiong, Ziyue Yang, Bolin Ni, Jingcheng Hu, Ruihang Li, Miaosen Zhang, Chen Li, Jia Ning, Ruizhe Wang, Zheng Zhang, Shuguang Liu, Han Hu, Peng Cheng
Tech Report
Tech Report  •   Code

ImageBrush: Learning Visual In-Context Instructions for Exemplar-Based Image Manipulation
2018 Model Compression, Computer Vision, Natural Language Processing, Neural Architecture Search

Yasheng Sun, Yifan Yang, Houwen Peng*, Yifei Shen, Yuqing Yang, Han Hu, Lili Qiu, Hideki Koike
NeurIPS 2023
Paper

TinyCLIP: CLIP Distillation via Affinity Mimicking and Weight Inheritance
2025 Foundation Models, Computer Vision, Multimodal Learning, Visual Tracking

Kan Wu, Houwen Peng*, Zhenghong Zhou, Bin Xiao, Mengchen Liu, Lu Yuan, Hong Xuan, Zhenghong Zhou, Xi Chen, Xinggang Wang, Hongyang Chao, Han Hu
ICCV 2023
Paper  •   Code

Attentive Mask CLIP
2025 Foundation Models, Natural Language Processing, Efficient AI

Yifan Yang, Weiquan Huang, Yixuan Wei, Houwen Peng*, Xinyang Jiang, Huiqiang Jiang, Fangyun Wei, Yin Wang, Han Hu, Lili Qiu, Yuqing Yang
ICCV 2023
Paper  •   Code

Exploring Lightweight Hierarchical Vision Transformers for Efficient Visual Tracking
2025 Model Compression, Neural Architecture Search, Transformer Architecture Selected

Ben Kang, Xin Chen, Dong Wang, Houwen Peng*, Huchuan Lu
ICCV 2023
Paper  •   Code

EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention
2017 Computer Vision, Large Language Models, Visual Tracking, Model Compression Selected

Xinyu Liu, Houwen Peng*, Ningxin Zheng, Yuqing Yang, Han Hu, Yixuan Yuan
CVPR 2023
Paper  •   Code

SeqTrack: Sequence to Sequence Learning for Visual Object Tracking
2017 Model Compression, Natural Language Processing, Multimodal Learning Selected

Xin Chen, Houwen Peng*, Dong Wang, Huchuan Lu, Han Hu
CVPR 2023
Paper  •   Code

iCLIP: Bridging Image Classification and Contrastive Language-Image Pre-Training for Visual Recognition
2018 Foundation Models, Efficient AI

Yixuan Wei, Yue Cao, Zheng Zhang, Houwen Peng, Zhuliang Yao, Zhenda Xie, Han Hu, Baining Guo
CVPR 2023
Paper

PointNeXt: Revisiting PointNet++ with Improved Training and Scaling Strategies
2016 Multimodal Learning, Efficient AI, Model Compression

Guocheng Qian, Yuchen Li, Houwen Peng*, Jinjie Mai, Hasan Abed Al Kader Hammoud, Mohamed Elhoseiny, Mohamed Elhoseiny, Bernard Ghanem*
NeurIPS 2022
Paper  •   Code

Expanding Language-Image Pretrained Models for General Video Recognition
2020 Large Language Models, Computer Vision, Visual Tracking

Bolin Ni, Houwen Peng*, Minghao Chen, Songyang Zhang, Gaofeng Meng, Jianlong Fu, Shiming Xiang, Haibin Ling
ECCV 2022
Oral Presentation   •   Paper  •   Code  •   🤗 Hugging Face

TinyViT: Fast Pretraining Distillation for Small Vision Transformers
2023 Natural Language Processing, Efficient AI, Foundation Models, Large Language Models

Kan Wu, Jinnian Zhang, Houwen Peng*, Mengchen Liu, Bin Xiao, Jianlong Fu, Lu Yuan
ECCV 2022
Paper  •   Code

MiniViT: Compressing Vision Transformers with Weight Multiplexing
2022 Visual Tracking, Foundation Models, Transformer Architecture, Model Compression

Jinnian Zhang, Houwen Peng*, Kan Wu, Mengchen Liu, Bin Xiao, Jianlong Fu, Lu Yuan
CVPR 2022
Paper  •   Code

Cyclic Differentiable Architecture Search
2016 Multimodal Learning, Visual Tracking, Computer Vision

Hongyuan Yu, Houwen Peng*, Yan Huang, Hao Du, Jianlong Fu, Liang Wang, Haibin Ling
TPAMI 2022
Paper  •   Code

Multi-Scale 2D Temporal Adjacent Networks for Moment Localization with Natural Language
2016 Visual Tracking, Natural Language Processing Selected

Songyang Zhang, Houwen Peng*, Jianlong Fu, Yijuan Lu, Jiebo Luo
TPAMI 2021
Paper  •   Code

AutoFormerV2: Searching the Search Space of Vision Transformer
2022 Neural Architecture Search, Computer Vision

Minghao Chen, Kan Wu, Bolin Ni, Houwen Peng*, Bei Liu, Jianlong Fu, Hongyang Chao, Haibin Ling
NeurIPS 2021
Paper  •   Code

Probing Inter-modality: Visual Parsing with Self-Attention for Vision-and-Language Pre-training
2023 Neural Architecture Search, Large Language Models Selected

Hongwei Xue, Yupan Huang, Bei Liu, Houwen Peng, Jianlong Fu, Houqiang Li, Jiebo Luo
NeurIPS 2021
Paper

Learning to Track Objects from Unlabled Videos
2022 Large Language Models, Model Compression Selected

Jilai Zheng, Chao Ma, Houwen Peng, Xiaokang Yang
ICCV 2021
Paper  •   Code

Rethinking and Improving Relative Position Encoding for Vision Transformer
2020 Large Language Models, Model Compression, Transformer Architecture

Kan Wu, Houwen Peng*, Minghao Chen, Jianlong Fu, Hongyang Chao
ICCV 2021
Paper  •   Code

AutoFormer: Searching Transformers for Visual Recognition
2023 Neural Architecture Search, Computer Vision, Multimodal Learning, Foundation Models

Minghao Chen, Houwen Peng*, Jianlong Fu, Haibin Ling
ICCV 2021
Paper  •   Code

Learning Spatio-Temporal Transformer for Visual Tracking
2015 Multimodal Learning, Visual Tracking, Natural Language Processing, Large Language Models

Bin Yan, Houwen Peng*, Jianlong Fu, Dong Wang, Huchuan Lu
ICCV 2021
Paper  •   Code

LightTrack: Finding Lightweight Neural Networks for Object Tracking via One-Shot Architecture Search
2025 Transformer Architecture, Multimodal Learning, Foundation Models, Natural Language Processing Selected

Bin Yan†, Houwen Peng†, Kan Wu†, Dong Wang, Jianlong Fu, Huchuan Lu
CVPR 2021
Paper  •   Code

One-Shot Neural Ensemble Architecture Search by Diversity-Guided Search Space Shrinking
2023 Efficient AI, Computer Vision Selected

Minghao Chen, Jianlong Fu, Haibin Ling
CVPR 2021
Paper  •   Code

Cream of the Crop: Distilling Prioritized Paths For One-Shot Neural Architecture Search
2020 Multimodal Learning, Foundation Models

Houwen Peng, Hao Du, Hongyuan Yu, Qi Li, Jing Liao, Jianlong Fu
NeurIPS 2020
Paper  •   Code

Ocean: Object-aware Anchor-free Tracking
2024 Transformer Architecture, Large Language Models

Zhipeng Zhang, Houwen Peng*, Jianlong Fu, Bing Li, Weiming Hu
ECCV 2020
Paper  •   Code

A Transductive Approach for Semi-Supervised Video Object Segmentation
2023 Efficient AI, Foundation Models

Yizhuo Zhang†, Zhirong Wu†, Houwen Peng, Stephen Lin
CVPR 2020
Paper  •   Code

Learning 2D Temporal Localization Networks for Moment Localization with Natural Language
2018 Computer Vision, Model Compression, Neural Architecture Search, Multimodal Learning

Songyang Zhang, Houwen Peng*, Jianlong Fu, Jiebo Luo
AAAI 2020
Paper  •   Code

Deeper and Wider Siamese Networks for Real-time Visual Tracking
2024 Efficient AI, Computer Vision, Large Language Models, Multimodal Learning Selected

Zhipeng Zhang, Houwen Peng*
CVPR 2019
Oral Presentation   •   Paper  •   Code

AI Coach: Deep Human Pose Estimation and Analysis for Personalized Athletic Training Assistance
2024 Multimodal Learning, Neural Architecture Search, Computer Vision

Jianbo Wang, Kai Qiu, Houwen Peng, Jianlong Fu, Jianke Zhu
ACM Multimedia 2019
Paper

Multi-view Multi-instance Learning based on Joint Sparse Representation and Multi-view Dictionary Learning
2019 Natural Language Processing, Large Language Models

Bing Li, Weihua Xiong, Houwen Peng, Weiming Hu, Stephen J. Maybank
TPAMI 2017
Paper

Illumination Estimation based on Bilayer Sparse Coding
2017 Natural Language Processing, Neural Architecture Search Selected

Bing Li, Weihua Xiong, Weiming Hu, Houwen Peng
CVPR 2013
Paper

Salient Object Detection via Structured Matrix Decomposition
2016 Foundation Models, Computer Vision, Large Language Models, Efficient AI

Houwen Peng, Haibin Ling, Bing Li, Weihua Xiong, Weiming Hu, Stephen J. Maybank
TPAMI 2017
Featured Paper   •   Paper  •   Code  •   Project Webpage (Updated)

Predicting Image Memorability by Multi-view Adaptive Regression
2021 Neural Architecture Search, Large Language Models, Computer Vision, Natural Language Processing Selected

Houwen Peng, Kai Li, Bing Li, Haibin Ling, Weihua Xiong, Weiming Hu
ACM Multimedia 2015
Paper

RGBD Salient Object Detection: A Benchmark and Algorithms
2023 Computer Vision, Efficient AI, Multimodal Learning

Houwen Peng, Bing Li, Weihua Xiong, Weiming Hu, Rongrong Ji
ECCV 2014
Paper  •   Code  •   Project Webpage

Salient Object Detection via Low-rank and Structured Sparse Matrix Decomposition
2021 Computer Vision, Foundation Models

Houwen Peng, Bing Li, Rongrong Ji, Weiming Hu, Weihua Xiong
AAAI 2013
Oral Presentation   •   Paper  •   Code  •   Project Webpage (Updated)

Group

Current and Former Interns/Collaborators