CE23 - Intelligence artificielle

Real-Time Analysis of Dynamic LiDAR 3D Point Clouds – READY3D

Submission summary

Autonomous driving is set to have a profound impact throughout our entire society in the near future. However, several scientific challenges remain unanswered, such as the automated analysis of dynamic 3D point clouds from moving vehicles at a precision and speed compatible with a fully-autonomous
system. In this project, our aim is to develop algorithms for the precise semantic and instance segmentation of mobile LiDAR scans, which can operate in real-time. This would be achieved by generalizing to dynamic data a state-of-the-art superpoint-based approach, which has been successful for static point clouds analysis.

Several interesting research questions at the interface between deep learning and functional optimization arise from considering such a problem. The results of these investigations would play a key role in increasing the efficiency and precision of dynamic 3D data analysis. Breakthroughs regarding these questions could have important other applications in fields such as computer vision, machine learning, and remote sensing. Finally, an open-source dataset would be released, incorporating the constraints and challenges of autonomous driving.

Project coordinator

Monsieur Loic Landrieu (Loic landrieu)

The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.

Partner

IGN Loic landrieu
Laboratoire d'Informatique Fondamentale et Appliquée de Tours (LIFAT)

Help of the ANR 194,400 euros
Beginning and duration of the scientific project: December 2019 - 42 Months

Useful links

Explorez notre base de projets financés

 

 

ANR makes available its datasets on funded projects, click here to find more.

Sign up for the latest news:
Subscribe to our newsletter