CuVLER: Enhanced Unsupervised Object Discoveries through Exhaustive Self-Supervised Transformers

1 code implementation • 12 Mar 2024 • Shahaf Arica, Or Rubin, Sapir Gershov, Shlomi Laufer

In this paper, we introduce VoteCut, an innovative method for unsupervised object discovery that leverages feature representations from multiple self-supervised models.

Clustering graph partitioning +4