Wen-Hsuan Chu

I am a Robotics Ph.D. at Prof. Katerina Fragkiakdaki's lab, at Carnegie Mellon University. I'm interested in the application of Computer Vision to Robotics and Embodied AI, such as performing 2D and 3D tracking at different granularities such as objects, parts, and points, and how this information can be used for robot manipulation or the learning of dynamics models.

Prior to this, I completed my M.S. at Carnegie Mellon University, in The Robotics Institute, advised by Prof. Kris Kitani, on machine learning and computer vision. Before this, I did a B.S. in Electrical Engineering at National Taiwan University.

CV Email
profile photo

Publications

Zero-Shot Open-Vocabulary Tracking with Large Pretrained Models

Under Submission

Wen-Hsuan Chu, Adam W. Harley, Pavel Tokmakov, Achal Dave, Leonidas Guibas, Katerina Fragkiadaki

project page paper
Tracking Any Point in a Video with Region-Aware Point Trajectory Transformer

Under submission

Wen-Hsuan Chu, Yizhou Zhao, Jie Li, Rares Ambrus, Adam W. Harley, Katerina Fragkiadaki

Neural Batch Sampling with Reinforcement Learning for Semi-Supervised Anomaly Detection

ECCV 2020

Wen-Hsuan Chu, Kris M. Kitani

paper
Spot and Learn: A Maximum-Entropy Patch Sampler for Few-Shot Image Classificatio

CVPR 2019

Wen-Hsuan Chu, Yu-Jhe Li, Jing-Cheng Chang, Yu-Chiang Frank Wang

paper
Learning Semantics-Guided Visual Attention for Few-shot Image Classification

ICIP 2018

Wen-Hsuan Chu, Yu-Chiang Frank Wang

paper