Supporting large-scale image recognition with out-of-domain samples
This work addresses the challenge of large-scale image recognition with out-of-domain samples, specifically for landmark image tasks, and is incremental as it builds on existing methods with novel re-ranking and filtering techniques.
The authors tackled the problem of labeling and ranking landmark images by developing an efficient end-to-end method that uses convolutional neural networks with an additive angular margin loss and re-ranks predictions using similarity to out-of-domain images, achieving first place in the 2020 Google Landmark Recognition challenge.
This article presents an efficient end-to-end method to perform instance-level recognition employed to the task of labeling and ranking landmark images. In a first step, we embed images in a high dimensional feature space using convolutional neural networks trained with an additive angular margin loss and classify images using visual similarity. We then efficiently re-rank predictions and filter noise utilizing similarity to out-of-domain images. Using this approach we achieved the 1st place in the 2020 edition of the Google Landmark Recognition challenge.