|
| 1 | +@article{chen2025cometokens, |
| 2 | + title = {Co-Me: Confidence-Guided Token Merging for Visual Geometric Transformers}, |
| 3 | + author = {Chen, Yutian and Qiu, Yuheng and Li, Ruogu and Agha, Ali and Omidshafiei, Shayegan and Patrikar, Jay and Scherer, Sebastian}, |
| 4 | + year = {2025}, |
| 5 | + url = {https://arxiv.org/abs/2511.14751}, |
| 6 | + journal = {arXiv preprint arXiv:2511.14751}, |
| 7 | + abstract = {We propose Confidence-Guided Token Merging (Co-Me), an acceleration mechanism for visual geometric transformers without retraining or finetuning the base model. Co-Me distilled a light-weight confidence predictor to rank tokens by uncertainty and selectively merge low-confidence ones, effectively reducing computation while maintaining spatial coverage. Compared to similarity-based merging or pruning, the confidence signal in Co-Me reliably indicates regions emphasized by the transformer, enabling substantial acceleration without degrading performance. Co-Me applies seamlessly to various multi-view and streaming visual geometric transformers, achieving speedups that scale with sequence length. When applied to VGGT and MapAnything, Co-Me achieves up to 11.3x and 7.2x speedup, making visual geometric transformers practical for real-time 3D perception and reconstruction.} |
| 8 | +} |
1 | 9 | @misc{yu2025unified, |
2 | 10 | title = {Unified Spherical Frontend: Learning Rotation-Equivariant Representations of Spherical Images from Any Camera}, |
3 | 11 | shorttitle = {Unified Spherical Frontend}, |
@@ -667,7 +675,7 @@ @article{moon2023time-optimal |
667 | 675 | } |
668 | 676 | @article{qiu2023airimu, |
669 | 677 | title = {{AirIMU: Learning} Uncertainty Propagation for Inertial Odometry}, |
670 | | - author = {Qiu, Yuheng and Wang, Chen and Zhou, Xunfei and Xia, Youjie and Scherer, Sebastian}, |
| 678 | + author = {Qiu, Yuheng and Wang, Chen and Xu, Can and Chen, Yutian and Zhou, Xunfei and Xia, Youjie and Scherer, Sebastian}, |
671 | 679 | year = {2023}, |
672 | 680 | journal = {arXiv preprint arXiv:2310.04874}, |
673 | 681 | url = {https://arxiv.org/pdf/2310.04874} |
@@ -1391,7 +1399,7 @@ @article{fang20203d-siamrpn |
1391 | 1399 | doi = {10.1109/JSEN.2020.3033034}, |
1392 | 1400 | url = {https://arxiv.org/pdf/2108.05630} |
1393 | 1401 | } |
1394 | | -@inproceedings{ho2020``provably, |
| 1402 | +@inproceedings{ho2020provably, |
1395 | 1403 | title = {``{{Provably Safe}}'' in the {{Wild}}: {{Testing Control Barrier Functions}} on a {{Vision-Based Quadrotor}} in an {{Outdoor Environment}}}, |
1396 | 1404 | author = {Ho, Cherie and Shih, Katherine and Grover, Jaskaran and Liu, Changliu and Scherer, Sebastian}, |
1397 | 1405 | year = {2020}, |
|
0 commit comments