About Me

I am a fourth-year PhD student in EECS at UC Merced under the advisement of Prof. Ming-Hsuan Yang. Before coming to UC Merced in 2016, I recieved my M.S. degree in the Ming Hsieh Department of Electrical Engineering from Univeristy of Southern California in 2016, and my B.S. degree in the Department of Electrical Engineering from National Taiwan University.

My research interests include Computer Vision and Machine Learning, particularly in video understanding and cross-modality modeling. Here is my CV.


  • library_books 05/2019: Google Internship, MTV
  • library_books 04/2019: One paper is accepted to ICIP 2019
  • library_books 02/2019: One paper is accepted to CVPR 2019
  • library_books 12/2018: Google Internship, SNV
  • library_books 07/2018: Two papers are accepted to ECCV 2018 (1 oral)
  • library_books 05/2018: Nvidia Internship, Santa Clara Audio-Visual Modeling
  • library_books 07/2017: One paper is accepted to ICCV 2017
  • library_books 05/2017: Google Internship, MTV Video Highlights / Video Summarization
  • library_books 08/2016: Start my PhD at UC Merced


Self-supervised Audio Spatialization with Correspondence Classifier
Yu-Ding Lu, Hsin-Ying Lee, Hung-Yu Tseng, Ming-Hsuan Yang
ICIP 2019
Mode Seeking Generative Adversarial Networks for Diverse Image Synthesis
Qi Mao*, Hsin-Ying Lee*, Hung-Yu Tseng*, Siwei Ma, and Ming-Hsuan Yang
CVPR 2019
Diverse Image-to-Image Translation via Disentangled Representations
Hsin-Ying Lee*, Hung-Yu Tseng*, Jia-Bin Huang, Maneesh Kumar Singh, and Ming-Hsuan Yang
ECCV 2018 (oral)
Sub-GAN: An Unsupervised Generative Model via Subspaces.
Jie Liang, Jufeng Yang, Hsin-Ying Lee, Kai Wang, and Ming-Hsuan Yang
ECCV 2018
Video Highlights Using Retention Stats
This project is done during the intership at Google. We explore a Youtube statistics - Retention Stats as an auxiliary training signal for video highlights extraction. Retention stats is a Youtube metric that takes multiple behaviors into consideration including when do audiences start, when do they quit, which parts are viewed multiple times and which parts are skipped frequently. We focus on sports highlights since they often share more consistent consensus among audience than other categories.
Hsin-Ying Lee, Min-Hsuan Tsai, Zheng Sun, and Weilong Yang
Unsupervised Representation Learning by Sorting Sequences
This work proposes a self-supervised sequence sorting task to learn image representation using unlabeled videos.
Hsin-Ying Lee, Jia-Bin Huang, Maneesh Kumar Singh, and Ming-Hsuan Yang
ICCV 2017
Graph-Based Logic Bit Slicing for Datapath-Aware Placement
This work proposes a balanced edge-cover algorithm to completely slice any datapath, and an SA-based refinement process to exploit the bit-slice paths as an alignment guidance to placers.
Chau-Chin Huang, Bo-Qiao Lin, Hsin-Ying Lee, Yao-Wen Chang, Kuo-Sheng Wu, and Jun-Zhi Yang
DAC 2017
Soft-Segmentation Guided Object Motion Deblurring
This work aims to use a maximum a posterior formulation in which soft-segmentation is incorporated for object layer estimation to jointly estimate object segmentation and camera motion.
Jinshan Pan, Zhe Hu, Hsin-Ying Lee, and Ming-Hsuan Yang
CVPR 2016
Detailed-routability-driven analytical placement for mixed-size designs with technology and region constraints
This work aims to participate in the 2015 ISPD Blockage-Aware Detailed Routing-Driven Placement Contest, which targeted at detailed-routability-driven analytical placement for mixed-size design.
Chau-Chin Huang, Hsin-Ying Lee, Bo-Qiao Lin, Sheng-Wei Yang, Chin-Hao Chang, Szu-To Chen, Yao-Wen Chang
ICCAD 2015
Bio-inspired proximity discovery and synchronization for D2D communications
This work aims to propose a distributed mechanism for Device-to-Device (D2D) communication, which achieves proximity discovery and synchronization simultaneously
Shih-Lung Chao, Hsin-Ying Lee, Ching-Chun Chou, Hung-Yu Wei
IEEE communications letters, 2013


Invited talk

Learning Visual-Semantic Embeddings with Structured Constraints
Jan. 2017, Academia Sinica, Taiwan

Invited talk

Unsupervised Representation Learning by Sorting Sequences
Dec. 2017, Academia Sinica, Taiwan

Oral Presentation

Diverse Image-to-Image Translation via Disentangled Representations
Sep. 2018, Munich, Germany


Ph.D. in EECS, University of California, Merced, California, USA
Sep. 2016 - present
M.S. in Electrical Engineering, University of Southern California, Los Angeles, California, USA
Sep. 2015 - Aug. 2016
B.S. in Electrical Engineering, National Taiwan University, Taipei, Taiwan
Sep. 2010 - Jun. 2014