Authors: Jongmin Yu, Chen Bene Chi, Sebastiano Fichera, Paolo Paoletti, Devansh Mehta, Shan Luo

Published on: February 06, 2024

Impact Score: 8.4

Arxiv code: Arxiv:2402.04064

Summary

• What is new: A novel end-to-end method that uses multiple spatial and channel-wise attention blocks for road defect detection and segmentation.
• Why this is important: Difficulty in developing an instance segmentation method that can detect and segment multiple types of road defects simultaneously.
• What the research proposes: A new method incorporating spatial and channel-wise attention blocks to better understand and segment various road defects.
• Results: The method outperforms existing state-of-the-art methods in multi-class road defect detection and segmentation on a newly collected dataset.

Technical Details

Technological frameworks used: nan

Models used: Spatial and channel-wise attention blocks

Data used: Newly collected dataset annotated with nine road defect classes

Potential Impact

Companies in the autonomous road repair systems market could greatly benefit from the advancements in detection and segmentation capabilities.

Want to implement this idea in a business?

We have generated a startup concept here: PaveSense.