Abstract
We present the first work where re-identification ofthe Giant Sunfish (Mola alexandrini) is automated using computer vision and deep learning. We propose a pipeline that scores an mAP of 60.34% on a full rank of the novel TinyMola dataset which includes 31 IDs and 91 images. The method requires no domain-adaptation or training which makes it especially suited for low-budget or volunteer-based projects, like Match My Mola, as part of a human-in-the-loop model.
The pipeline includes segmentation, keypoint detection and description, keypoint matching, and ranking. The choice of feature descriptor has the largest impact on the performance and we show that the deep learning based SuperPoint descriptor greatly outperforms handcrafted descriptors like SIFT and RootSIFT independent of the segmentation level and matching method. Combining SuperPoint and the graph neural network based SuperGlue matching method produces the best results.
The pipeline includes segmentation, keypoint detection and description, keypoint matching, and ranking. The choice of feature descriptor has the largest impact on the performance and we show that the deep learning based SuperPoint descriptor greatly outperforms handcrafted descriptors like SIFT and RootSIFT independent of the segmentation level and matching method. Combining SuperPoint and the graph neural network based SuperGlue matching method produces the best results.
Original language | English |
---|---|
Journal | Proceedings of the Northern Lights Deep Learning Workshop |
Volume | 3 |
Number of pages | 9 |
ISSN | 2703-6928 |
DOIs | |
Publication status | Published - 28 Mar 2022 |
Event | Northern Lights Deep Learning Workshop - Tromsø, Norway Duration: 10 Jan 2022 → 12 Jan 2022 https://www.nldl.org/ |
Workshop
Workshop | Northern Lights Deep Learning Workshop |
---|---|
Country/Territory | Norway |
City | Tromsø |
Period | 10/01/2022 → 12/01/2022 |
Internet address |