Tips and Tricks for Webly-Supervised Fine-Grained Recognition: Learning from the WebFG 2020 Challenge

TL;DR

This paper reviews the WebFG 2020 challenge focused on webly-supervised fine-grained recognition, highlighting effective methods and surprising findings from top solutions to improve practical model training with web data.

Contribution

It compiles and analyzes the top solutions from the WebFG 2020 challenge, providing insights into effective strategies for webly-supervised fine-grained recognition.

Findings

Certain web data augmentation techniques significantly improve accuracy.

Some complex models did not outperform simpler approaches.

Effective use of web data can reduce reliance on labeled datasets.

Abstract

WebFG 2020 is an international challenge hosted by Nanjing University of Science and Technology, University of Edinburgh, Nanjing University, The University of Adelaide, Waseda University, etc. This challenge mainly pays attention to the webly-supervised fine-grained recognition problem. In the literature, existing deep learning methods highly rely on large-scale and high-quality labeled training data, which poses a limitation to their practicability and scalability in real world applications. In particular, for fine-grained recognition, a visual task that requires professional knowledge for labeling, the cost of acquiring labeled training data is quite high. It causes extreme difficulties to obtain a large amount of high-quality training data. Therefore, utilizing free web data to train fine-grained recognition models has attracted increasing attentions from researchers in the fine-grained community. This challenge expects participants to develop webly-supervised fine-grained recognition methods, which leverages web images in training fine-grained recognition models to ease the extreme dependence of deep learning methods on large-scale manually labeled datasets and to enhance their practicability and scalability. In this technical report, we have pulled together the top WebFG 2020 solutions of total 54 competing teams, and discuss what methods worked best across the set of winning teams, and what surprisingly did not help.