Hyun Oh Song

60 5th Ave, New York, NY 10011

I'm an associate professor in the Department of Computer Science and Engineering at Seoul National University. Previously, I was a research scientist at Google Research, in Mountain View, where I worked on machine learning research in Kevin Murphy's team. Before Google, I was a postdoctoral fellow in SAIL in the Computer Science Department at Stanford University.

I received my Ph.D. in Computer Science at UC Berkeley in 2014, where I worked with Trevor Darrell and Stefanie Jegelka. My graduate study was fully supported by Samsung Lee Kun Hee Scholarship Foundation for five years. In summer 2013, I spent time at LEAR, INRIA as a visiting student researcher. I did my B.S. at Hanyang University.

My research interests are in machine learning, combinatorial optimization, and algorithms. Broadly, I'm interested in solving challenging problems in artificial intelligence.

Email / Google Scholar / Github (mllab) / Github (personal) / LinkedIn

News

I'll serve as an Area chair at NeurIPS 2021, 2022, 2023, 2024, ICML 2023, 2024, and a Senior program committee member at AAAI 2022.
I'll serve as a virtual deep-dive session chair at NeurIPS 2022 on offline reinforcement learning.
I gave an invited talk at Samsung AI Forum 2022 on learning with combinatorial structures.
I founded an AI health care startup, DeepMetrics. We raised $1.1M VC seed funding in May 2021 and additional $1.1M government research funding in July 2021. We are actively hiring!
Tensorflow code for the suite of deep metric learning methods I worked on at Google is officially open sourced at TensorFlow Addons.
I'm moving to Seoul National University as an assistant professor starting Sep 2017.
I'll join Google Research as a research scientist starting July, 2016.
I'll serve as an Oral and Spotlight session chair at CVPR 2016.

For prospective students

For undergrads at SNU, you should take my deep learning class (M2177.0043) and machine learning class (4190.666). For math & stat, I recommend you to take these classes: 300.203A (881.007), M1407.000600 (881.008), 881.319, 326.211, 326.311, 3341.454. For CS, please take: M1522.0009, 4190.407.
It's best to first apply for undergraduate internship. Please contact (via email) at least one semester before your graduate application for the internship.

Selected publications

	LayerMerge: Neural Network Depth Compression through Layer Pruning and Merging Jinuk Kim, Marwa El Halabi, Mingi Ji, Hyun Oh Song International Conference on Machine Learning (ICML), 2024 paper / code / bibtex / project page / poster
	Training Greedy Policy for Proposal Batch Selection in Expensive Multi-Objective Combinatorial Optimization Deokjae Lee, Hyun Oh Song, Kyunghyun Cho International Conference on Machine Learning (ICML), 2024 paper / code / bibtex
	Compressed Context Memory For Online Language Model Interaction Jang-Hyun Kim, Junyoung Yeom, Sangdoo Yun, Hyun Oh Song International Conference on Learning Representations (ICLR), 2024 paper / code / bibtex / project page
	Direct Preference-based Policy Optimization without Reward Modeling Gaon An, Junhyeok Lee, Xingdong Zuo, Norio Kosaka, Kyung-Min Kim, Hyun Oh Song Neural Information Processing Systems (NeurIPS), 2023 paper / code / bibtex
	Neural Relation Graph: A Unified Framework for Identifying Label Noise and Outlier Data Jang-Hyun Kim, Sangdoo Yun, Hyun Oh Song Neural Information Processing Systems (NeurIPS), 2023 paper / code / bibtex
	Discovering Hierarchical Achievements in Reinforcement Learning via Contrastive Learning Seungyong Moon, Junyoung Yeom, Bumsoo Park, Hyun Oh Song Neural Information Processing Systems (NeurIPS), 2023 paper / code / bibtex
	Query-Efficient Black-Box Red Teaming via Bayesian Optimization Deokjae Lee, JunYeong Lee, Jung-Woo Ha, Jin-Hwa Kim, Sang-Woo Lee, Hwaran Lee, Hyun Oh Song Association for Computational Linguistics (ACL), 2023 paper / code / bibtex
	Efficient Latency-Aware CNN Depth Compression via Two-Stage Dynamic Programming Jinuk Kim, Yeonwoo Jeong, Deokjae Lee, Hyun Oh Song International Conference on Machine Learning (ICML), 2023 paper / code / bibtex
	Rethinking Value Function Learning for Generalization in Reinforcement Learning Seungyong Moon, JunYeong Lee, Hyun Oh Song Neural Information Processing Systems (NeurIPS), 2022 paper / code / bibtex
	Dataset Condensation via Efficient Synthetic-Data Parameterization Jang-Hyun Kim, Jinuk Kim, Seong Joon Oh, Sangdoo Yun, Hwanjun Song, Joonhyun Jeong, Jung-Woo Ha, Hyun Oh Song International Conference on Machine Learning (ICML), 2022 paper / code / bibtex
	Query-Efficient and Scalable Black-Box Adversarial Attacks on Discrete Sequential Data via Bayesian Optimization Deokjae Lee, Seungyong Moon, Junhyeok Lee, Hyun Oh Song International Conference on Machine Learning (ICML), 2022 paper / code / bibtex
	Optimal channel selection with discrete QCQP Yeonwoo Jeong, Deokjae Lee, Gaon An, Changyong Son, Hyun Oh Song International Conference on Artificial Intelligence and Statistics (AISTATS), 2022 paper / code / bibtex
	Preemptive Image Robustification for Protecting Users against Man-in-the-Middle Adversarial Attacks Seungyong Moon, Gaon An, Hyun Oh Song AAAI Conference on Artificial Intelligence (AAAI), 2022 paper / code / bibtex
	Uncertainty-Based Offline Reinforcement Learning with Diversified Q-Ensemble Gaon An, Seungyong Moon, Jang-Hyun Kim, Hyun Oh Song Neural Information Processing Systems (NeurIPS), 2021 paper / code / bibtex
	Co-Mixup: Saliency Guided Joint Mixup with Supermodular Diversity Jang-Hyun Kim, Wonho Choo, Hosan Jeong, Hyun Oh Song International Conference on Learning Representations (ICLR), 2021 Oral presentation (53/2997=1.7%) paper / supp / code / bibtex
	Puzzle Mix: Exploiting Saliency and Local Statistics for Optimal Mixup Jang-Hyun Kim, Wonho Choo, Hyun Oh Song International Conference on Machine Learning (ICML), 2020 paper / supp / code / bibtex / talk video
	End-to-End Efficient Representation Learning via Cascading Combinatorial Optimization Yeonwoo Jeong, Yoonsung Kim, Hyun Oh Song IEEE Computer Vision and Pattern Recognition (CVPR), 2019 paper / supp / code / bibtex
	Parsimonious Black-Box Adversarial Attacks via Efficient Combinatorial Optimization Seungyong Moon, Gaon An, Hyun Oh Song International Conference on Machine Learning (ICML), 2019 Long talk (159/3424=4.6%) paper / supp / code / bibtex / talk video
	Learning Discrete and Continuous Factors of Data via Alternating Disentanglement Yeonwoo Jeong, Hyun Oh Song International Conference on Machine Learning (ICML), 2019 paper / supp / code / bibtex / talk video
	EMI: Exploration with Mutual Information Hyoungseok Kim, Jaekyeom Kim, Yeonwoo Jeong, Sergey Levine, Hyun Oh Song International Conference on Machine Learning (ICML), 2019 Long talk (159/3424=4.6%) paper / supp / code / bibtex / talk video
	Efficient end-to-end learning for quantizable representations Yeonwoo Jeong, Hyun Oh Song International Conference on Machine Learning (ICML), 2018 Long talk (213/2473=8.6%) paper / code / bibtex / talk video
	Deep Metric Learning via Facility Location Hyun Oh Song, Stefanie Jegelka, Vivek Rathod, Kevin Murphy IEEE Computer Vision and Pattern Recognition (CVPR), 2017 Spotlight presentation (144/2680=5.3%) paper / code / bibtex / talk video
	Learning Transferrable Representations for Unsupervised Domain Adaptation Ozan Sener, Hyun Oh Song, Ashutosh Saxena, Silvio Savarese Neural Information Processing Systems (NIPS), 2016 paper / bibtex
	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 (123/2145=5.7%) paper / code / dataset / bibtex / talk video
	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
	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
	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
	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
	Discriminatively Activated Sparselets Hyun Oh Song, Ross Girshick, Trevor Darrell International Conference on Machine Learning (ICML), 2013 Long talk (143/1204=11.8%) paper / slide / poster / supp / bibtex
	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
	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