Abstract
Convolutional neural networks (CNN) have demonstrated remarkable performance, when the training and testing data are from the same distribution. However, such trained CNN models often largely degrade on testing data which is unseen and Out-Of-the-Distribution (OOD). To address this issue, we propose a novel “Decoupled-Mixup" method to train CNN models for OOD visual recognition. Different from previous work combining pairs of images homogeneously, our method decouples each image into discriminative and noise-prone regions, and then heterogeneously combine these regions of image pairs to train CNN models. Since the observation is that noise-prone regions such as textural and clutter background are adverse to the generalization ability of CNN models during training, we enhance features from discriminative regions and suppress noise-prone ones when combining an image pair. To further improves the generalization ability of trained models, we propose to disentangle discriminative and noise-prone regions in frequency-based and context-based fashions. Experiment results show the high generalization performance of our method on testing data that are composed of unseen contexts, where our method achieves 85.76% top-1 accuracy in Track-1 and 79.92% in Track-2 in NICO Challenge. The source code is available at https://github.com/HaozheLiu-ST/NICOChallenge-OOD-Classification.
| Original language | English (US) |
|---|---|
| Title of host publication | Computer Vision – ECCV 2022 Workshops, Proceedings |
| Editors | Leonid Karlinsky, Tomer Michaeli, Ko Nishino |
| Publisher | Springer Science and Business Media Deutschland GmbH |
| Pages | 451-464 |
| Number of pages | 14 |
| ISBN (Print) | 9783031250743 |
| DOIs | |
| State | Published - 2023 |
| Event | 17th European Conference on Computer Vision, ECCV 2022 - Tel Aviv, Israel Duration: Oct 23 2022 → Oct 27 2022 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 13806 LNCS |
| ISSN (Print) | 0302-9743 |
| ISSN (Electronic) | 1611-3349 |
Conference
| Conference | 17th European Conference on Computer Vision, ECCV 2022 |
|---|---|
| Country/Territory | Israel |
| City | Tel Aviv |
| Period | 10/23/22 → 10/27/22 |
Bibliographical note
Funding Information:Acknowledgements. We would like to thank the efforts of the NICO Challenge officials, who are committed to maintaining the fairness and openness of the competition. We also could not have undertaken this paper without efforts of every authors. This work was supported in part by the Key-Area Research and Development Program of Guangdong Province, China (No. 2018B010111001), National Key R&D Program of China (2018YFC2000702), in part by the Scientific and Technical Innovation 2030-“New Generation Artificial Intelligence” Project (No. 2020AAA0104100) and in part by the King Abdullah University of Science and Technology (KAUST) Office of Sponsored Research through the Visual Computing Center (VCC) funding.
Funding Information:
We would like to thank the efforts of the NICO Challenge officials, who are committed to maintaining the fairness and openness of the competition. We also could not have undertaken this paper without efforts of every authors. This work was supported in part by the Key-Area Research and Development Program of Guangdong Province, China (No. 2018B010111001), National Key R&D Program of China (2018YFC2000702), in part by the Scientific and Technical Innovation 2030“New Generation Artificial Intelligence” Project (No. 2020AAA0104100) and in part by the King Abdullah University of Science and Technology (KAUST) Office of Sponsored Research through the Visual Computing Center (VCC) funding.
Publisher Copyright:
© 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
ASJC Scopus subject areas
- Theoretical Computer Science
- General Computer Science