Poster · Sydney
G²TR: Generation-Guided Visual Token Reduction for Separate-Encoder Unified Multimodal Models
- When
- Tue 8 Dec 2026, 10:00–13:00 AEDT
- Where
- Hall 1-4, ICC Sydney
Exhibition Centre, Level 1 - Session
- Sydney Poster Session 1
- Authors
- Junxian Li, Kai Liu, Zizhong Ding, Zhixin Wang, Zhikai Chen, Renjing Pei, Yulun Zhang
- Organisations
- Shanghai Jiao Tong University, Huawei
- QRS classification
- Computer vision