Hyun Oh Song

I'm a research scientist at Google Research, in Mountain View, where I work on machine learning, computer vision, and deep learning. Before Google, I was a postdoctoral fellow in SAIL in the Computer Science Department at Stanford University. Before Stanford, I finished my Ph.D in Computer Science at UC Berkeley in late 2014. My graduate study was fully supported by Samsung Lee Kun Hee Scholarship Foundation (Now Samsung Scholarship Foundation) for five years. In summer 2013, I spent a wonderful time at LEAR, INRIA as a visiting student researcher.

My research interests are in machine learning, computer vision, and deep learning. Broadly, I'm interested in solving challenging problems in artificial intelligence. Also, I'm an invited reviewer for NIPS, ICML, SIGGRAPH, TPAMI, IJCV, RSS, ICCV, ECCV, CVPR, CVIU, and ICRA.

Email / LinkedIn / Thesis / Demo video 1 / Demo video 2 / Demo video 3 / Github



News

  • I'll join Google Research as a research scientist starting July, 2016.
  • I'll serve as an Oral and Spotlight session chair at CVPR16.
  • I'm giving a guest lecture on Convolutional Neural Networks for CS231A at Stanford on May 18.
  • Code and dataset on Deep metric learning are now available on Github.
  • I'm co-organizing an ONR workshop on Structured Learning at Stanford.


Selected Publications
arxiv2016

Unsupervised Transductive Domain Adaptation
Ozan Sener, Hyun Oh Song, Ashutosh Saxena, Silvio Savarese
arXiv:1511.06452 [stat.ML; cs.LG], 2016
paper

cvpr2016

Deep Metric Learning via Lifted Structured Feature Embedding
Hyun Oh Song, Yu Xiang, Stefanie Jegelka, Silvio Savarese
IEEE Computer Vision and Pattern Recognition (CVPR), 2016
Spotlight presentation
paper / code / dataset / bibtex

Sparselet_TPAMI2014

Generalized Sparselet Models for Real-Time Multiclass Object Recognition
Hyun Oh Song, Ross Girshick, Stefan Zickler, Christopher Geyer, Pedro Felzenszwalb, Trevor Darrell
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2015
paper / code / Demo Video 1 / Demo Video 2 / bibtex

TASE2014

Learning to detect visual grasp affordance
Hyun Oh Song, Mario Fritz, Daniel Goehring, Trevor Darrell
IEEE Transactions on Automation Science and Engineering (TASE), 2015
paper / Demo Video / bibtex

OBOD_ICML14

Weakly-supervised Discovery of Visual Pattern Configurations
Hyun Oh Song, Yong Jae Lee, Stefanie Jegelka, Trevor Darrell
Neural Information Processing Systems (NIPS), 2014
paper / bibtex

OBOD_ICML14

On learning to localize objects with minimal supervision
Hyun Oh Song, Ross Girshick, Stefanie Jegelka, Julien Mairal, Zaid Harchaoui, Trevor Darrell
International Conference on Machine Learning (ICML), 2014
paper / Talk Video / code / bibtex

DAS_ICML13

Discriminatively Activated Sparselets
Hyun Oh Song*, Ross Girshick*, Trevor Darrell
International Conference on Machine Learning (ICML), 2013
Selected full oral presentation
paper / slide / poster / supp / bibtex

Bbank_ACMMM12

Detection Bank: An Object Detection Based Video Representation for Multimedia Event Recognition
Tim Althoff, Hyun Oh Song, Trevor Darrell
ACM Multimedia (ACMMM), 2012
paper / poster / bibtex

Sparselets_ECCV12

Sparselet Models for Efficient Multiclass Object Detection
Hyun Oh Song, Stefan Zickler, Tim Althoff, Ross Girshick, Mario Fritz, Christopher Geyer, Pedro Felzenszwalb, Trevor Darrell
European Conference on Computer Vision (ECCV), 2012
paper / Demo Video / poster / bibtex / code

Grasp_ICCV11

Visual Grasp Affordances From Appearance-Based Cues
Hyun Oh Song, Mario Fritz, Chunhui Gu, Trevor Darrell
ICCV Workshop on Challenges and Opportunities in Robot Perception, 2011
Best paper award
paper / bibtex


Demos
Demo_CVPR12

Real-Time Multiclass Object Recognition
Hyun Oh Song, Stefan Zickler, Ross Girshick, Christopher Geyer, Pedro Felzenszwalb, Trevor Darrell
Computer Vision and Pattern Recognition (CVPR), 2012


Awards
SLSF_Logo

Samsung Lee Kun Hee Foundation (now Samsung Scholarship)
5 year Ph.D. fellowship


Teaching
CS70

CS70 Discrete Mathematics and Probability Theory
GSI, UC Berkeley, 2012


Erdös = 4 (via three paths)

I like this website