Haoran Song

I am a Ph.D. candidate at the Robotics Institute, Hong Kong University of Science and Technology, where I work on robotics and machine learning under the supervision of Prof. Qifeng Chen and Prof. Michael Yu Wang.

I was a visiting scholar at the Robotics, Perception and Learning Lab, KTH Royal Institute of Technology, Sweden in 2018, working with Prof. Danica Kragic. Before that, I graduated from Harbin Institute of Technology in 2016 with a bachelor's degree in Flight Vehicle Design and Engineering.

Email  /  CV  /  Google Scholar  /  Github

profile photo
Research

My research lies in motion forecasting and planning. Early in my Ph.D. study, much of my research is about robotics planning under physical interaction. Since last year, I stepped on the journey towards autonomous driving and injected traditional planning ideas into learning-based methods. Now I'm dedicated to working on trajectory prediction and motion planning for autonomous driving. My first-authored works are listed below.

Ongoing projects: I'm currently developing robust and interpretable algorithms on multi-modal prediction under urban scenarios.

PiP architecture PiP: Planning-informed Trajectory Prediction for Autonomous Driving
Haoran Song, Wenchao Ding, Yuxuan Chen, Shaojie Shen, Michael Yu Wang, Qifeng Chen
European Conference on Computer Vision (ECCV 2020), Glasgow, Scotland, UK

Inform the multi-agent future prediction with ego vehicle's planning in a novel planning-prediction-coupled pipeline.

project page  /  arXiv  /  bibtex
PiP architecture Multi-Object Rearrangement with Monte Carlo Tree Search:A Case Study on Planar Nonprehensile Sorting
Haoran Song*, Joshua A. Haustein*, Weihao Yuan, Kaiyu Hang, Michael Yu Wang, Danica Kragic, Johannes A. Stork
IEEE International Conference on Intelligent Robots and Systems (IROS 2020), Las Vegas, USA

Learn policy from MCTS and employ it to reinforce MCTS, to address sorting task under multi-contact physics.

project page  /  arXiv  /  bibtex
Herding visualization Herding by Caging
Haoran Song*, Anastasiia Varava*, Danica Kragic, Michael Yu Wang, Florian T. Pokorny, Kaiyu Hang
Under review at Autonomous Robots (AURO), 2019

Apply computational geometry and topology in planning a set of mobile robots to address the interesting 'herding' task.

Cover image of Science Robotics Perching and resting - A paradigm for UAV maneuvering with modularized landing gears
Kaiyu Hang*, Ximin Lyu*, Haoran Song*, Johannes A Stork*, Aaron M Dollar, Danica Kragic, Fu Zhang (* denotes equal contribution)
Science Robotics, Cover Article in March 2019

Drones that perch like birds could go on much longer flights!

Media Reports: MIT Technology Review, Nature Electronics, National Public Radio, etc.

paper  /  video  /  bibtex

Grasp planning results. Fingertip surface optimization for robust grasping on contact primitives
Haoran Song, Michael Yu Wang, Kaiyu Hang
IEEE Robotics and Automation Letters (RA-L), 2018 &
IEEE International Conference on Robotics and Automation (ICRA 2018), Brisbane, Australia

Rethink grasp planning from geometry matching rather than traditional point contact modeling.

paper  /  bibtex
Other Projects

These include other projects and issued patents.

Grasp planning results. Reinforcement Learning in Topology-based Representation for Human Body Movement with Whole Arm Manipulation
Weihao Yuan, Kaiyu Hang, Haoran Song, Danica Kragic, Michael Y. Wang, Johannes A. Stork
IEEE International Conference on Robotics and Automation (ICRA 2019), Montreal, Canada

paper  /  bibtex  /  code

key points involved in the patent Robotic fingertip design and grasping on contact primitives
Kaiyu Hang, Haoran Song
US Patent App. 16/192,169, 2019

patent info  /  bibtex

Service
Reviewer for IEEE Transactions on Robotics (T-RO)
Reviewer for IEEE Robotics and Automation Letters (RA-L)
Reviewer for IEEE International Conference on Robotics and Automation (ICRA)
Reviewer for IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

Modified from Dr. Jon Barron's design of the website.