HotSpotter - Patterned Species Instance Recognition
This work addresses the need for efficient and non-species-specific instance recognition in wildlife monitoring, though it is incremental as it builds on existing techniques like Local Naive Bayes Nearest Neighbor.
The paper tackles the problem of identifying individual animals from images by presenting HotSpotter, a fast and accurate algorithm that matches query images against a labeled database, achieving more accurate results than existing methods and processing each query in just a few seconds on databases of over 1000 images.
We present HotSpotter, a fast, accurate algorithm for identifying individual animals against a labeled database. It is not species specific and has been applied to Grevy's and plains zebras, giraffes, leopards, and lionfish. We describe two approaches, both based on extracting and matching keypoints or "hotspots". The first tests each new query image sequentially against each database image, generating a score for each database image in isolation, and ranking the results. The second, building on recent techniques for instance recognition, matches the query image against the database using a fast nearest neighbor search. It uses a competitive scoring mechanism derived from the Local Naive Bayes Nearest Neighbor algorithm recently proposed for category recognition. We demonstrate results on databases of more than 1000 images, producing more accurate matches than published methods and matching each query image in just a few seconds.