Hyeonwoo Noh

Computer Vision Lab
Department of Computer Science and Engineering
POSTECH (Pohang University of Science and Technology)
Advisor : Prof. Bohyung Han

News

  • [April, 2019] We released dataset and code for the paper "Transfer Learning via Unsupervised Task Discovery for Visual Question Answering".   [paper] [code]

  • [March, 2019] Our paper entitled "Exploiting Hierarchy for Learning and Transfer in KL-regularized RL" is now available on arXiv. This work is done during my internship at DeepMind   [paper]

  • [Februrary, 2019] Our paper entitled "Transfer Learning via Unsupervised Task Discovery for Visual Question Answering" is accepted to CVPR 2019.   [paper]

  • [December, 2018] Our paper entitled "Transfer Learning via Unsupervised Task Discovery for Visual Question Answering" is now available on arXiv.   [paper]

  • [July, 2018] I will intern at Google DeepMind, advised by Yee Whye Teh. 2018/7 - 2018/12.

  • [June, 2018] Our paper "Neural Program Synthesis from Diverse Demonstration Videos" is accepted to ICML 2018. [paper][project page] This work is done during my visiting at CLVR lab, USC

  • [March, 2018] DELF (DEep Local Features), introduced in our paper "Large-Scale Image Retrieval with Attentive Deep Local Features", achieved state-of-the-art results in a CVPR'18 image retrieval paper: Radenovic et al., "Revisiting Oxford and Paris: Large-Scale Image Retrieval Benchmarking".

  • [September, 2017] Our paper "Regularizing Deep Neural Networks by Noise: Its Interpretation and Optimization" is accepted to NIPS 2017. [paper][poster][video]

  • [July, 2017] Our paper "Large-Scale Image Retrieval with Attentive Deep Local Features" is accepted to ICCV 2017. We empirically show that local descriptor based image retrieval system still ourperforms the global descriptor based one in a large-scale instance retrieval setting and propose an attentive deep local features. This work is done during my internship at Google. [paper][code]

  • [October, 2016] I was selected as one of the Microsoft Research Asia Fellows.

  • [June, 2016] I did my internship at Google (LA), 2016/6 - 2016/12.

  • [June, 2016] Our paper entitled "Training Recurrent Answering Units with Joint Loss Minimization for VQA" is now available on arXiv.   [paper] [project page]

  • [March, 2016] Our paper "Image Question Answering using Convolutional Neural Network with Dynamic Parameter Prediction" has been accepted as oral presentation to CVPR 2016   [paper] [project page]

  • [September, 2015] Our paper "Decoupled Deep Neural Network for Semi-supervised Semantic Segmentation" has been accepted as spotlight presentation to NIPS 2015   [paper] [project page].

  • [September, 2015] Our paper "Learning Deconvolution Network for Semantic Segmentation" has been accepted to ICCV 2015   [paper] [project page].

  • [Febrary, 2015] I joined Computer Vision Lab, POSTECH as a Ph.D. student

Education

Industrial Experience

  • Research Intern. DeepMind., London, United Kingdom (July, 2018 - December, 2018)

  • Research Intern. Google Inc., Los Angeles, California, USA (June, 2016 - December, 2016)

  • Software Engineer. DMC R&D Center, Samsung Electronics, Korea (July, 2013 - January, 2015)

Publications

*Dhruva Tirumala, *Hyeonwoo Noh, Alexandre Galashov, Leonard Hasenclever,
Arun Ahuja, Greg Wayne, Razvan Pascanu, Yee Whye Teh, Nicolas Heess
In Arxiv (*Equal contribution)
[arXiv preprint]
 
Hyeonwoo Noh, Taehoon Kim, Jonghwan Mun, Bohyung Han,
To appear in CVPR 2019
[arXiv preprint][code]
 
*Shao-Hua Sun, *Hyeonwoo Noh, Sriram Somasundaram, and Joseph Lim,
In ICML 2018 (*Equal contribution)
[paper][project page]
 
Hyeonwoo Noh, Tackgeun You, Jonghwan Mun, and Bohyung Han,
In NIPS 2017
[arXiv preprint][paper][poster][video]
 
Hyeonwoo Noh, Andre Araujo, Jack Sim, Tobias Weyand, Bohyung Han,
In ICCV 2017
[arXiv preprint][paper][code]
 

*Seunghoon Hong, *Hyeonwoo Noh, Bohyung Han,
In NIPS 2015 (*Equal contribution)
[arXiv preprint] [paper] [project page]

 
Hyeonwoo Noh, Seunghoon Hong, Bohyung Han,
In ICCV 2015
[arXiv preprint] [paper] [project page]
 
Hyeonwoo Noh, Bohyung Han,
In ACM Multimedia(MM) 2012
[pdf]
 

Contact Information