# Google Landmark Recognition 2020 Competition Third Place Solution

arXiv: https://arxiv.org/abs/2010.05350 · HTML: https://mlsanity.com/paper/2010.05350/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 11 Oct 2020.

## Built on by (most cited first, 9 of 9)

- [1st Place Solution in Google Universal Images Embedding](https://mlsanity.com/paper/2210.08473/) (2022)
- [EViT: Privacy-Preserving Image Retrieval via Encrypted Vision Transformer in Cloud Computing](https://mlsanity.com/paper/2208.14657/) (2022)
- [Efficient large-scale image retrieval with deep feature orthogonality and Hybrid-Swin-Transformers](https://mlsanity.com/paper/2110.03786/) (2021)
- [ScaleFace: Uncertainty-aware Deep Metric Learning](https://mlsanity.com/paper/2209.01880/) (2022)
- [3rd Place Solution to Google Landmark Recognition Competition 2021](https://mlsanity.com/paper/2110.02794/) (2021)
- [6th Place Solution to Google Universal Image Embedding](https://mlsanity.com/paper/2210.09377/) (2022)
- [Efficient and Discriminative Image Feature Extraction for Universal Image Retrieval](https://mlsanity.com/paper/2409.13513/) (2024)
- [Universal Embedding Function for Traffic Classification via QUIC Domain Recognition Pretraining: A Transfer Learning Success](https://mlsanity.com/paper/2502.12930/) (2025)
- [Google Landmark Retrieval 2021 Competition Third Place Solution](https://mlsanity.com/paper/2110.04619/) (2021)

## Abstract (arXiv)

We present our third place solution to the Google Landmark Recognition 2020 competition. It is an ensemble of global features only Sub-center ArcFace models. We introduce dynamic margins for ArcFace loss, a family of tune-able margin functions of class size, designed to deal with the extreme imbalance in GLDv2 dataset. Progressive finetuning and careful postprocessing are also key to the solution. Our two submissions scored 0.6344 and 0.6289 on private leaderboard, both ranking third place out of 736 teams.