PerceptionDLM Uses Diffusion Language Models for Region-Level Vision

PerceptionDLM applies multimodal diffusion language models to parallel region perception tasks in images and video. No benchmark numbers or architecture details are available from the source.