Can I read 3rd Place Solution for Google Universal Image Embedding on EtoBox?

3rd Place Solution for Google Universal Image Embedding by Aoki, Nobuaki; Namba, Yasumasa is a scholarly article available to read on EtoBox.

What is 3rd Place Solution for Google Universal Image Embedding about?

This paper presents the 3rd place solution to the Google Universal Image Embedding Competition on Kaggle. We use ViT-H/14 from OpenCLIP for the backbone of ArcFace, and trained in 2 stage. 1st stage is done with freezed backbone, and 2nd stage is whole model training. We achieve 0.692 mean Precision @5 on private leaderboard. Code available at https://github.com/YasumasaNamba/google-universal-image-embedding

Author

Aoki, Nobuaki; Namba, Yasumasa

Published

2022

Language

EN