TY  - RPRT
TI  - DDCL: Deep Dual Competitive Learning: A Differentiable End-to-End Framework for Unsupervised Prototype-Based Representation Learning
AU  - Giansalvo Cirrincione
PY  - 2026
UR  - https://arxiv.org/abs/2604.01740
ID  - 2604.01740
ER  -