Bandwidth-adaptive Cloud-Assisted 360-Degree 3D Perception for Autonomous Vehicles
Faisal Hawlader, Rui Meireles, Gamal Elghazaly, Ana Aguiar, Raphaël Frank · 2026
Read Bandwidth-adaptive Cloud-Assisted 360-Degree 3D Perception for Autonomous Vehicles by Faisal Hawlader, Rui Meireles, Gamal Elghazaly, Ana Aguiar, Raphaël Frank (2026) 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-2602-23871.txt.
Description
Faisal Hawladera and four co-authors present a cloud-assisted perception system that gives self-driving cars a full 360-degree three-dimensional view while adapting to available bandwidth. The 2026 paper addresses the clash between heavy onboard compute and strict latency limits, which bites hardest in busy urban scenes.
Transformer models fuse multi-camera data into a bird's-eye view for three-dimensional object detection. Work splits between car and cloud, with the split point and feature quantization chosen on the fly; feature clipping and compression cut network load before transmission. Real-world tests show a large latency cut versus a purely onboard system. Under varying bandwidth, the adaptive choice of split and quantization improves accuracy over fixed settings at the same latency, keeping perception timely when networks fluctuate.
Catalog record
| CREATOR | Faisal Hawlader, Rui Meireles, Gamal Elghazaly, Ana Aguiar, Raphaël Frank |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 69 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-23871