PerceptionLM: Open-Access Data and Models for Detailed Visual Understanding

Cho, Jang Hyun, Madotto, Andrea, Mavroudi, Effrosyni · arXiv · 2025