3rd Place Solution to "Google Landmark Retrieval 2020"
This is an incremental improvement for image retrieval in computer vision, specifically for landmark recognition tasks.
The paper tackled the Google Landmark Retrieval 2020 challenge by focusing on data cleaning and metric learning models, achieving 3rd place in the competition.
Image retrieval is a fundamental problem in computer vision. This paper presents our 3rd place detailed solution to the Google Landmark Retrieval 2020 challenge. We focus on the exploration of data cleaning and models with metric learning. We use a data cleaning strategy based on embedding clustering. Besides, we employ a data augmentation method called Corner-Cutmix, which improves the model's ability to recognize multi-scale and occluded landmark images. We show in detail the ablation experiments and results of our method.