#artificial intelligence
Jun 2026
Rethinking Object-Centric Representations for Video Dynamics Modeling
Unified Slots (Unified Slots), an unsupervised framework for learning robust and disentangled object-centric representations from videos, is introduced, achieving state-of-the-art performance in unsupervised object-centric video decomposition and tracking.
Amaury Wei, Ismail Nejjar, Olga Fink
· arXiv.org · 1 citation