Huazhe(Harry) Xu

About Me

I am a third year Ph.D student in Berkeley AI Research (BAIR) advised by Prof. Trevor Darrell. I received my bachelor degree from Tsinghua University. I have also spent wonderful time at research labs of Facebook AI, University of Toronto and International Computer Science Institute enjoying the collaboration with Prof. Tengyu Ma, Dr. Yuandong Tian, Prof. Jiashi Feng, Prof. Sergey Levine, Prof. Sanja Fidler and Prof. Raquel Urtasun.

My research focuses on modelling th dynamics of the world, leveraging human priors for policy learning, and further enabling learning algorithms to learn in a sample-efficient manner. I am also interestded in solving complex video games/real applications with deep learning and reinforcement learning.

I am an amateur pianist and actively looking for potential collaborations (both music-wise or research-wise!). If you do reinforcement learning or computer vision projects or you play the piano, violin or cello, etc, feel free to contact me for some potential projects or some fun!

Education

Aug. 2012 - Jul. 2016 , Department of Electronic Engineering, Tsinghua University,

Balchlor of Engineering, GPA: 93/100, ranking: 5/238. Average of Math and Math-Related Courses: 95.4/100.

Aug. 2014 - Dec. 2014, School of Electrical and Computer Engineering, University of Toronto,

Exchange Student, GPA: 4.0/4.0.

July. 2015 - Sept. 2015, Aug. 2016 - now , Department of Electrical and Computer Engineering, University of California, Berkeley,

Visiting Researcher, PhD Student.

Selected Research Projects

Modeling Visual Dynamics

Learning Diverse and Natural Behaviors from Only Observations,

Joint work w/ J. Xu T. Darrell

Ongoing project [pdf] [code]

Object-Centric Video Prediction,

Joint work w/ B. Chen, M. Yang, Y. Gao, T. Darrell

Ongoing project [pdf] [code]

Disentangling Propagation and Generation for Video Prediction,

Joint work w/ H. Gao, Q. Cai, R. Wang, F. Yu, T. Darrell

To appear in ICCV'19 [pdf] [code]

Human priors and dynamics models for efficient policy learning

Model-based Multiagent Reinforcement Learning for Fast Adaptation,

Joint work w/ Y. Luo, B. Chen, T. Darrell

Ongoing Project [pdf] [code]

Learning Self-Correctable Policies and Value Functions from Demonstrations with Negative Sampling,

Joint work w/ Y. Luo and T. Ma

Submitted [pdf] [code]

Algorithmic Framework for Model-based Deep Reinforcement learning with Theoretical Guarantees,

Joint work w/ Y. Luo, Y. Li, Y. Tian, T. Darrell, T. Ma

published at International Conference on Learning Representations 2019 [pdf] [code] [website]

Reinforcement Learning from Imperfect Demonstrations,

Joint work w/ Y. Gao, J. Lin, F. Yu, S. Levine and T. Darrell

Neurips Deep RL Symposium [pdf] [code]

Learning policies for Real-world Applications

End-to-End Learning of Driving Models from Large-scale Video Datasets,

Joint work w/ Y. Gao, F. Yu and T. Darrell

published at Conference on Computer Vision and Pattern Recognition 2017 (oral) [pdf] [code]

Modular Architecture for StarCraft II with Deep Reinforcement Learning,

Joint work w/ D. Lee, H. Tang, J. Zhang, T. Darrell and P. Abbeel

published at AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment 2018 [pdf] [code]

Past Projects

Natural Language and Object Retrieval,

Joint work w/ R. Hu, M. Rohrbach, J. Feng, K. Saenko and T. Darrell,

published at Conference on Computer Vision and Pattern Recognition 2016 (oral) [pdf] [code]

Automobile Visual Taste Ranking,

Joint work w/ S. Fidler, R. Urtasun,

2015 Fall , University of Toronto,

Publications and Manuscripts

• Y. Luo, H. Xu, T. Ma, Learning Self-Correctable Policies and Value Functions from Demonstrations with Negative Sampling,, Arxiv Preprint, . [pdf]

• H. Gao^*, H. Xu^*, Q. Cai, R. Wang, F. Yu, T. Darrell, Disentangling Propagation and Generation for Video Prediction, To Appear in ICCV'19, . [pdf]

• H. Tang^*, D. Lee^*, J. Zhang, H. Xu, T. Darrell, P. Abbeel, Modular Architecture for StarCraft II with Deep Reinforcement Learning. The 14th AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment (AIIDE'18). [pdf]

• Y. Luo^*, H. Xu^*, Y. Li, Y. Tian, T. Darrell, T. Ma, Algorithmic Framework for Model-based Reinforcement Learning with Theoretical Guarantees .ICLR'19 (also appeared as ICMLW'18, NeurIPSW'18). [pdf]

• Y. Gao^*, H. Xu^*, F. Yu, S. Levine, T. Darrell, Reinforcement Learning from Imperfect Demonstrations. in NIPS Deep RL Syposium. [pdf]

• H. Xu^*, Y. Gao^*, F. Yu, T. Darrell, End-to-end Learning of Driving Models from Large-scale Video Datasets. in CVPR 2017 (oral). [pdf]

• R. Hu, H. Xu, M. Rohrbach, J. Feng, K. Saenko, T. Darrell, Natural Language Object Retrieval. in CVPR 2016 (oral). [pdf]

• Tian Xie, Qian Han, Huazhe Xu, Zihao Qi, Wenqian Shen. A Low-Complexity Linear Precoding Scheme Based on SOR Method for Massive MIMO Systems. Vehicular Technology Conference (VTC Spring), 2015 IEEE 81st , vol., no., pp.1-5, 11-14 May 2015. [pdf]

Honors and Awards

• EECS Excellence Award, UC Berkeley, 2016

• Comprehensive Excellent Scholarship, Tsinghua Univ, 2013

• Academic Excellent Scholarship, Tsinghua, Univ, 2014

• Academic Excellent Scholarship, Tsinghua, Univ, 2015

News!

• July, 2019: One paper accepted by ICCV'19.

• May, 2019: One paper accepted by ICLR'19.

• Nov 19, 2018, Our Starcraft 2 Project is covered by Synced (机器之心) and Import AI.

• Dec 18, 2017: Giving a talk at Berkeley DeepDrive Workshop.

• Oct 31, 2017: Giving a talk at MIT vision seminar in Boston.

• June 1, 2017: Giving a talk in at Tsinghua University in Beijing.

• Our paper 'End-to-end Learning of Driving Model from Largescale Dataset' appeared at 'New_era' Wechat Media.

• Interviewed by a 'Leiphone' reporter about BAIR Blog.

Service

• Reviewer of ICLR, Neurips, CVPR, ICML’19

• Reviewer of ACCV, Neurips, ICML, CVPR’18

• Reviewer of NIPS'17 Intelligent Transportation Workshop

• Reviewer of IEEE Transactions on Multimedia’ 16, 17

• Editor in BAIR Blog Editorial Board, manager of BAIR facebook handle.

Miscellaneous

• Alain De Botton: The Art of Travel (I traveled less often after reading this :p)

• Introduction to Image Understanding (A lovely Course from Prof. Sanja Fidler)

• Scriabin Etude op. 2 No. 1 in C sharp minor played by Horowitz (The piece of music I listened most often.)

• Franz Schubert Sonata D960 (Kafka on The Shore novel by Murakami brought Schubert's D850 to me when I was 16.)

Attempts

Piano Project

I am now learning to play the piano(again!) with pianist Anne Rainwater. Here are some pieces I enjoy playing recently:

Three Romances for Violin and Piano Op. 22, Clara Schumann

Phatasiestucke, Robert Schumann

Piano Sonata No. 3 in C major, Op. 2, No. 3, Ludwig van Beethoven

Impromptu in B-flat Major D. 935, No. 3 (Op. 142), Franz Schubert

French Suites, BWV 817, Johann Sebastian Bach

Prelude and Fugue No.2, Johann Sebastian Bach

4 Improptus, D899, Franz Schubert

kinderszenen op. 15, Robert Schumann

Waltz in D-flat major, Op. 64, No. 1, Valse du petit chien, Frédéric Chopin

Grande valse brillante in E-flat major, Op. 18, Frédéric Chopin

Tempest Sonata, Ludwig van Beethoven

Reading Project

I enjoy reading sci-fi novels and love stories. Recently finished Isaac Asimov's The Greater Foundation Series and The Gods Themselves.

Tennis

I also enjoy playing tennis. If you are visiting Berkeley and want to play, ping me on email.

Who searched for me!