TY  - RPRT
TI  - Featurising Pixels from Dynamic 3D Scenes with Linear In-Context Learners
AU  - Nikita Araslanov
AU  - Martin Sundermeyer
AU  - Hidenobu Matsuki
AU  - David Joseph Tan
AU  - Federico Tombari
PY  - 2026
UR  - https://arxiv.org/abs/2604.26488
ID  - 2604.26488
ER  -