TY  - RPRT
TI  - T-REN: Learning Text-Aligned Region Tokens Improves Dense Vision-Language Alignment and Scalability
AU  - Savya Khosla
AU  - Sethuraman T V
AU  - Aryan Chadha
AU  - Alex Schwing
AU  - Derek Hoiem
PY  - 2026
UR  - https://arxiv.org/abs/2604.18573
ID  - 2604.18573
ER  -