Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox

Picture for Seong Joon Oh

Rethinking Spatial Dimensions of Vision Transformers


Mar 30, 2021
Byeongho Heo, Sangdoo Yun, Dongyoon Han, Sanghyuk Chun, Junsuk Choe, Seong Joon Oh

* 10 pages, 5 figures 

  Access Paper or Ask Questions

Probabilistic Embeddings for Cross-Modal Retrieval


Jan 13, 2021
Sanghyuk Chun, Seong Joon Oh, Rafael Sampaio de Rezende, Yannis Kalantidis, Diane Larlus

* 17 pages, 20 figures 

  Access Paper or Ask Questions

Re-labeling ImageNet: from Single to Multi-Labels, from Global to Localized Labels


Jan 13, 2021
Sangdoo Yun, Seong Joon Oh, Byeongho Heo, Dongyoon Han, Junsuk Choe, Sanghyuk Chun

* 15 pages, 10 figures, tech report 

  Access Paper or Ask Questions

VideoMix: Rethinking Data Augmentation for Video Classification


Dec 07, 2020
Sangdoo Yun, Seong Joon Oh, Byeongho Heo, Dongyoon Han, Jinhyung Kim

* 15 pages 

  Access Paper or Ask Questions

Evaluation for Weakly Supervised Object Localization: Protocol, Metrics, and Datasets


Jul 08, 2020
Junsuk Choe, Seong Joon Oh, Sanghyuk Chun, Zeynep Akata, Hyunjung Shim

* TPAMI submission. First two authors contributed equally. This is a journal extension of our CVPR 2020 paper arXiv:2001.07437. Code: https://github.com/clovaai/wsolevaluation 

  Access Paper or Ask Questions

Slowing Down the Weight Norm Increase in Momentum-based Optimizers


Jun 15, 2020
Byeongho Heo, Sanghyuk Chun, Seong Joon Oh, Dongyoon Han, Sangdoo Yun, Youngjung Uh, Jung-Woo Ha

* 16 pages, 9 figures. First two authors contributed equally 

  Access Paper or Ask Questions

An Empirical Evaluation on Robustness and Uncertainty of Regularization Methods


Mar 09, 2020
Sanghyuk Chun, Seong Joon Oh, Sangdoo Yun, Dongyoon Han, Junsuk Choe, Youngjoon Yoo

* Accepted at ICML 2019 Workshop on Uncertainty and Robustness in Deep Learning. 7 pages, 1 figure 

  Access Paper or Ask Questions

Reliable Fidelity and Diversity Metrics for Generative Models


Feb 23, 2020
Muhammad Ferjad Naeem, Seong Joon Oh, Youngjung Uh, Yunjey Choi, Jaejun Yoo

* First two authors have contributed equally 

  Access Paper or Ask Questions

Evaluating Weakly Supervised Object Localization Methods Right


Jan 21, 2020
Junsuk Choe, Seong Joon Oh, Seungho Lee, Sanghyuk Chun, Zeynep Akata, Hyunjung Shim


  Access Paper or Ask Questions

On Recognizing Texts of Arbitrary Shapes with 2D Self-Attention


Oct 10, 2019
Junyeop Lee, Sungrae Park, Jeonghun Baek, Seong Joon Oh, Seonghyeon Kim, Hwalsuk Lee


  Access Paper or Ask Questions

Learning De-biased Representations with Biased Representations


Oct 07, 2019
Hyojin Bahng, Sanghyuk Chun, Sangdoo Yun, Jaegul Choo, Seong Joon Oh

* 17 pages, 11 figures 

  Access Paper or Ask Questions

CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features


May 13, 2019
Sangdoo Yun, Dongyoon Han, Seong Joon Oh, Sanghyuk Chun, Junsuk Choe, Youngjoon Yoo

* 14 pages, 5 figures 

  Access Paper or Ask Questions

What is wrong with scene text recognition model comparisons? dataset and model analysis


Apr 03, 2019
Jeonghun Baek, Geewook Kim, Junyeop Lee, Sungrae Park, Dongyoon Han, Sangdoo Yun, Seong Joon Oh, Hwalsuk Lee

* Our code will be publicly available. (https://github.com/clovaai/deep-text-recognition-benchmark

  Access Paper or Ask Questions

Person Recognition in Personal Photo Collections


Oct 20, 2018
Seong Joon Oh, Rodrigo Benenson, Mario Fritz, Bernt Schiele

* 18 pages, 20 figures; to appear in IEEE Transactions on Pattern Analysis and Machine Intelligence 

  Access Paper or Ask Questions

Modeling Uncertainty with Hedged Instance Embedding


Oct 19, 2018
Seong Joon Oh, Kevin Murphy, Jiyan Pan, Joseph Roth, Florian Schroff, Andrew Gallagher

* 15 pages, 10 figures 

  Access Paper or Ask Questions

Sequential Attacks on Agents for Long-Term Adversarial Goals


Jul 05, 2018
Edgar Tretschk, Seong Joon Oh, Mario Fritz


  Access Paper or Ask Questions

Understanding and Controlling User Linkability in Decentralized Learning


May 15, 2018
Tribhuvanesh Orekondy, Seong Joon Oh, Bernt Schiele, Mario Fritz


  Access Paper or Ask Questions

Natural and Effective Obfuscation by Head Inpainting


Mar 16, 2018
Qianru Sun, Liqian Ma, Seong Joon Oh, Luc Van Gool, Bernt Schiele, Mario Fritz

* To appear in CVPR 2018 

  Access Paper or Ask Questions

Towards Reverse-Engineering Black-Box Neural Networks


Feb 14, 2018
Seong Joon Oh, Max Augustin, Bernt Schiele, Mario Fritz

* 20 pages, 12 figures, to appear at ICLR'18. Code: https://goo.gl/MbYfsv 

  Access Paper or Ask Questions

Adversarial Image Perturbation for Privacy Protection -- A Game Theory Perspective


Jul 26, 2017
Seong Joon Oh, Mario Fritz, Bernt Schiele

* To appear at ICCV'17 

  Access Paper or Ask Questions

Exploiting saliency for object segmentation from image level labels


Jul 14, 2017
Seong Joon Oh, Rodrigo Benenson, Anna Khoreva, Zeynep Akata, Mario Fritz, Bernt Schiele

* CVPR 2017 

  Access Paper or Ask Questions

Generating Descriptions with Grounded and Co-Referenced People


Apr 05, 2017
Anna Rohrbach, Marcus Rohrbach, Siyu Tang, Seong Joon Oh, Bernt Schiele

* Accepted to CVPR 2017 

  Access Paper or Ask Questions

Faceless Person Recognition; Privacy Implications in Social Media


Jul 28, 2016
Seong Joon Oh, Rodrigo Benenson, Mario Fritz, Bernt Schiele

* Accepted to ECCV'16 

  Access Paper or Ask Questions