CrowdVLM-R1: Expanding R1 Ability to Vision Language Model for Crowd Counting using Fuzzy Group Relative Policy Reward Paper • 2504.03724 • Published Mar 31 • 1
