Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Publisher
Springer Science and Business Media Deutschland GmbH
Citation
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol.15480 LNCS, pp.419-435
Adapting deep learning networks for point cloud data recognition in self-driving vehicles faces challenges due to the variability in datasets and sensor technologies, emphasizing the need for adaptive techniques to maintain accuracy across different conditions. In this paper, we introduce the 3D Adaptive Structural Convolution Network (3D-ASCN), a cutting-edge framework for 3D point cloud recognition. It combines 3D convolution kernels, a structural tree structure, and adaptive neighborhood sampling for effective geometric feature extraction. This method obtains domain-invariant features and demonstrates robust, adaptable performance on a variety of point cloud datasets, ensuring compatibility across diverse sensor configurations without the need for parameter adjustments. This highlights its potential to significantly enhance the reliability and efficiency of self-driving vehicle technology.
We would like to express sincere gratitude to Beomsik Cho and Seonghoon Ryoo (Ajou University) for their invaluable assistance with the conceptualization and experiments in this work. Their contributions greatly supported the development of this research. This work was supported in 2024 by Korea National Police Agency (KNPA) under the project \\u201CDevelopment of autonomous driving patrol service for active prevention and response to traffic accidents\\u201D (RS-2024-00403630), Institute of Information communications Technology Planning & Evaluation (IITP) under the Artificial Intelligence Convergence Innovation Human Resources Development (IITP-2023-No. RS-2023-00255968) grant funded by the Korea government (MSIT), and the BK21 FOUR program of the National Research Foundation Korea funded by the Ministry of Education (NRF5199991014091).