Showing cs.CVShow all
2 papers · 1 filter
cs.CV2026
Self-Supervised Representation Learning via Hyperspherical Density Shaping
Esteban Rodríguez-Betancourt, Edgar Casasola-Murillo
Modern self-supervised representation learning methods often relies on empirical heuristics that are not theoretically grounded. In this study we propose HyDeS, a theoretically gro…
cs.CV2026
Hypersolid: Emergent Vision Representations via Short-Range Repulsion
Esteban Rodríguez-Betancourt, Edgar Casasola-Murillo
A central problem in self-supervised learning is preventing representation collapse. Most methods avoid it through global mechanisms, such as contrastive expansion, variance constr…