1st Place Solution for ECCV 2022 OOD-CV Challenge Image Classification Track

TL;DR

This paper presents a top-performing solution for out-of-distribution image classification that combines domain generalization pre-training with test-time domain adaptation using noisy label learning, achieving first place in the ECCV 2022 OOD-CV Challenge.

Contribution

It introduces a novel Mask-Level Copy-Paste data augmentation and a Label-Periodically-Updated DivideMix method for effective test-time noisy label learning.

Findings

Achieved first place in the ECCV 2022 OOD-CV Challenge.

Effective domain generalization with Mask-Level Copy-Paste augmentation.

Successful test-time adaptation using noisy label learning.

Abstract

OOD-CV challenge is an out-of-distribution generalization task. In this challenge, our core solution can be summarized as that Noisy Label Learning Is A Strong Test-Time Domain Adaptation Optimizer. Briefly speaking, our main pipeline can be divided into two stages, a pre-training stage for domain generalization and a test-time training stage for domain adaptation. We only exploit labeled source data in the pre-training stage and only exploit unlabeled target data in the test-time training stage. In the pre-training stage, we propose a simple yet effective Mask-Level Copy-Paste data augmentation strategy to enhance out-of-distribution generalization ability so as to resist shape, pose, context, texture, occlusion, and weather domain shifts in this challenge. In the test-time training stage, we use the pre-trained model to assign noisy label for the unlabeled target data, and propose a Label-Periodically-Updated DivideMix method for noisy label learning. After integrating Test-Time Augmentation and Model Ensemble strategies, our solution ranks the first place on the Image Classification Leaderboard of the OOD-CV Challenge. Code will be released in https://github.com/hikvision-research/OOD-CV.