Real-Time Obstacle Avoidance for a Mobile Robot Using CNN-Based Sensor Fusion
1 st Lamiaa H. Zain · 2025
Read Real-Time Obstacle Avoidance for a Mobile Robot Using CNN-Based Sensor Fusion by 1 st Lamiaa H. Zain (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-2509-08095.txt.
Description
A mobile robot that cannot dodge obstacles cannot work outside the lab. Lamiaa H. Zain trains neural networks for that skill in a 2025 research paper written alone. Three convolutional networks were built end to end. They take synchronized color and depth images from an Intel RealSense camera and output steering commands directly.
Offline tests favored the model called NetConEmb, with the lowest angular error. A lighter model, NetEmb, used about a quarter fewer parameters and converged faster, with accuracy close behind. In live navigation, NetConEmb succeeded in every trial, in both familiar and new environments. The lighter models managed only the familiar setting. The results show what real-time sensor fusion costs and buys for small robots that must react without a human at the controls.
Catalog record
| CREATOR | 1 st Lamiaa H. Zain |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 18 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-08095