2nd Place Solution to Google Landmark Retrieval 2021
This work addresses landmark retrieval for computer vision applications, but it is incremental as it builds on existing methods with specific optimizations for a competition.
The paper tackled the Google Landmark Retrieval 2021 competition by developing a solution that improved retrieval accuracy, achieving a mean average precision of 0.52995 at rank 100 on the private leaderboard.
This paper presents the 2nd place solution to the Google Landmark Retrieval 2021 Competition on Kaggle. The solution is based on a baseline with training tricks from person re-identification, a continent-aware sampling strategy is presented to select training images according to their country tags and a Landmark-Country aware reranking is proposed for the retrieval task. With these contributions, we achieve 0.52995 mAP@100 on private leaderboard. Code available at https://github.com/WesleyZhang1991/Google_Landmark_Retrieval_2021_2nd_Place_Solution