STAMP: S calable T ask- A nd M odel-agnostic Collaborative P erception
Xiangbo Gao, Runsheng Xu, Jiachen Li, Ziran Wang, Zhiwen Fan, Zhengzhong Tu · 2025
Read STAMP: S calable T ask- A nd M odel-agnostic Collaborative P erception by Xiangbo Gao, Runsheng Xu, Jiachen Li, Ziran Wang, Zhiwen Fan, Zhengzhong Tu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-18616.txt.
Description
Xiangbo Gao and five co-authors build a way for self-driving cars with different brains to share what they see, in this 2025 paper. A single car's sensors are blocked by trucks, weather, and distance. Fleets could pool their views, but cars from different makers run different perception models whose internal formats do not match.
Their pipeline, STAMP, sidesteps the clash with translator pairs. Each car converts its bird's-eye-view features into a shared protocol format on the way out, and back into its private format on the way in. The translators are light, cheap to run, and reveal nothing about the model behind them, which protects trade secrets. On simulated and real roads, the pooled system matched or beat leading models at a fraction of the computing cost. It is presented as a step toward fleets of mixed brands perceiving as one.
Catalog record
| CREATOR | Xiangbo Gao, Runsheng Xu, Jiachen Li, Ziran Wang, Zhiwen Fan, Zhengzhong Tu |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-18616