1 paper · 1 filter
Renping Zhou, Zanlin Ni, Tianyi Chen +6
Recently, Masked Diffusion Models (MDMs) have shown promising potential across vision, language, and cross-modal generation. However, a notable discrepancy exists between their tra…