Timezone: »
Unsupervised learning has recently made exceptional progress because of the development of more effective contrastive learning methods. However, CNNs are prone to depend on low-level features that humans deem non-semantic. This dependency has been conjectured to induce a lack of robustness to image perturbations or domain shift. In this paper, we show that by generating carefully designed negative samples, contrastive learning can learn more robust representations with less dependence on such features. Contrastive learning utilizes positive pairs which preserve semantic information while perturbing superficial features in the training images. Similarly, we propose to generate negative samples in a reversed way, where only the superfluous instead of the semantic features are preserved. We develop two methods, texture-based and patch-based augmentations, to generate negative samples. These samples achieve better generalization, especially under out-of-domain settings. We also analyze our method and the generated texture-based samples, showing that texture features are indispensable in classifying particular ImageNet classes and especially finer classes. We also show that the model bias between texture and shape features favors them differently under different test settings.
Author Information
Songwei Ge (University of Maryland, College Park)
Shlok Mishra (University of Maryland, College Park)
Chun-Liang Li (Google)
Haohan Wang (Carnegie Mellon University)
David Jacobs (University of Maryland, USA)
More from the Same Authors
-
2021 : Improving Model Compatibility of Generative Adversarial Networks by Boundary Calibration »
Si-An Chen · Chun-Liang Li · Hsuan-Tien Lin -
2021 : Improving Model Compatibility of Generative Adversarial Networks by Boundary Calibration »
Si-An Chen · Chun-Liang Li · Hsuan-Tien Lin -
2022 Poster: Autoregressive Perturbations for Data Poisoning »
Pedro Sandoval-Segura · Vasu Singla · Jonas Geiping · Micah Goldblum · Tom Goldstein · David Jacobs -
2021 Poster: Shift Invariance Can Reduce Adversarial Robustness »
Vasu Singla · Songwei Ge · Basri Ronen · David Jacobs -
2021 Poster: A Unified View of cGANs with and without Classifiers »
Si-An Chen · Chun-Liang Li · Hsuan-Tien Lin -
2021 Poster: Object-aware Contrastive Learning for Debiased Scene Representation »
Sangwoo Mo · Hyunwoo Kang · Kihyuk Sohn · Chun-Liang Li · Jinwoo Shin -
2020 : Creative Sketch Generation »
Songwei Ge -
2019 : Poster Session 2 »
Mayur Saxena · Nicholas Frosst · Vivien Cabannes · Gene Kogan · Austin Dill · Anurag Sarkar · Joel Ruben Antony Moniz · Vibert Thio · Scott Sievert · Lia Coleman · Frederik De Bleser · Brian Quanz · Jonathon Kereliuk · Panos Achlioptas · Mohamed Elhoseiny · Songwei Ge · Aidan Gomez · Jamie Brew -
2019 : Coffee Break & Poster Session 1 »
Yan Zhang · Jonathon Hare · Adam Prugel-Bennett · Po Leung · Patrick Flaherty · Pitchaya Wiratchotisatian · Alessandro Epasto · Silvio Lattanzi · Sergei Vassilvitskii · Morteza Zadimoghaddam · Theja Tulabandhula · Fabian Fuchs · Adam Kosiorek · Ingmar Posner · William Hang · Anna Goldie · Sujith Ravi · Azalia Mirhoseini · Yuwen Xiong · Mengye Ren · Renjie Liao · Raquel Urtasun · Haici Zhang · Michele Borassi · Shengda Luo · Andrew Trapp · Geoffroy Dubourg-Felonneau · Yasmeen Kussad · Christopher Bender · Manzil Zaheer · Junier Oliva · Michał Stypułkowski · Maciej Zieba · Austin Dill · Chun-Liang Li · Songwei Ge · Eunsu Kang · Oiwi Parker Jones · Kelvin Ka Wing Wong · Joshua Payne · Yang Li · Azade Nazi · Erkut Erdem · Aykut Erdem · Kevin O'Connor · Juan J Garcia · Maciej Zamorski · Jan Chorowski · Deeksha Sinha · Harry Clifford · John W Cassidy -
2019 : Posters »
Colin Graber · Yuan-Ting Hu · Tiantian Fang · Jessica Hamrick · Giorgio Giannone · John Co-Reyes · Boyang Deng · Eric Crawford · Andrea Dittadi · Peter Karkus · Matthew Dirks · Rakshit Trivedi · Sunny Raj · Javier Felip Leon · Harris Chan · Jan Chorowski · Jeff Orchard · Aleksandar Stanić · Adam Kortylewski · Ben Zinberg · Chenghui Zhou · Wei Sun · Vikash Mansinghka · Chun-Liang Li · Marco Cusumano-Towner -
2019 Poster: Learning Robust Global Representations by Penalizing Local Predictive Power »
Haohan Wang · Songwei Ge · Zachary Lipton · Eric Xing -
2019 Poster: The Convergence Rate of Neural Networks for Learned Functions of Different Frequencies »
Basri Ronen · David Jacobs · Yoni Kasten · Shira Kritchman -
2018 : Poster Session 1 »
Evan Casey · Colin A Raffel · Jonathan Simon · Juncheng Li · Robert Saunders · Petra Gemeinboeck · Eunsu Kang · Songwei Ge · Curtis Hawthorne · Anna Huang · Ting-Wei Su · Eric Chu · Memo Akten · Sonam Damani · Khyatti Gupta · Dilpreet Singh · Patrick Hutchings -
2018 Poster: Nonparametric Density Estimation under Adversarial Losses »
Shashank Singh · Ananya Uppal · Boyue Li · Chun-Liang Li · Manzil Zaheer · Barnabas Poczos -
2017 Poster: MMD GAN: Towards Deeper Understanding of Moment Matching Network »
Chun-Liang Li · Wei-Cheng Chang · Yu Cheng · Yiming Yang · Barnabas Poczos