T-REN (
Text-aligned
Region
Encoder
Network) is an efficient encoder that maps visual data to a compact set of text-aligned region-level representations (region tokens). It is built on top of a frozen
DINOv3 ViT-L/16 backbone and adds only 3.7% additional parameters.
Compared to patch-based vision-language backbones, T-REN yields stronger dense cross-modal understanding while significantly reducing token counts by more than 24x for images and 187x for videos.
1@misc{khosla2026tren,
2 title={T-REN: Learning Text-Aligned Region Tokens Improves Dense Vision-Language Alignment and Scalability},
3 author={Savya Khosla and Sethuraman T V and Aryan Chadha and Alexander Schwing and Derek Hoiem},
4 year={2026},
5 archivePrefix={arXiv},
6 primaryClass={cs.CV},
7}
