About this document
Memory-Augmented Self-Supervised Tracker by Jan Kristanto is a document available to read on EtoBox.
MAST is a self-supervised dense tracking model that uses a memory module to learn from past frames without human annotations. It outperforms other self-supervised baselines on DAVIS-2017 and YouTube-VOS benchmarks, achieving 15% and 17% higher mean Jaccard and F-measure scores respectively. Qualitative results also show MAST produces more accurate predictions over time compared to other self-supervised methods.
- Author
- Jan Kristanto
- Language
- EN