Authors: Nicolaj Schmid, Cornelius von Einem, Cesar Cadena, Roland Siegwart, Lorenz Hruby, Florian Tschopp

Published on: March 14, 2024

Impact Score: 7.8

Arxiv code: Arxiv:2403.09477

Summary

• What is new: Introduces VIRUS-NeRF, a novel approach using low-cost sensors for mapping environments by robots.
• Why this is important: The high cost and complexity of LiDAR and depth cameras for obstacle detection and path planning in robotics.
• What the research proposes: A system called VIRUS-NeRF that uses inexpensive ultrasonic and infrared sensors combined with Neural Radiance Fields for effective mapping.
• Results: Comparable mapping performance to LiDAR in small environments and a significant increase in training speed.

Technical Details

Technological frameworks used: Instant Neural Graphics Primitives with Multiresolution Hash Encoding (Instant-NGP)

Models used: Neural Radiance Fields (NeRF)

Data used: Depth measurements from ultrasonic and infrared sensors

Potential Impact

Robotics, automation, manufacturing, and warehousing companies could benefit, while high-cost sensor manufacturers might face disruption.

Want to implement this idea in a business?

We have generated a startup concept here: SentryPath.