Shohei Wakita
Papers
1
Total Citations
6
H-Index
1
About
Shohei Wakita is a researcher specializing in robotics, autonomous navigation, and sensor data compression, with a particular focus on laser-based mapping and self-localization. His most notable contribution is the development of the Laser Variational Autoencoder (Laser VAE), a novel approach that leverages deep generative models to compress laser-scan data for efficient map construction and accurate global self-localization. This work, published in 2018 and cited 6 times, addresses a critical challenge in robotics: reducing memory usage while maintaining localization precision. Unlike traditional methods that rely on handcrafted feature extractors tailored to specific environments like offices or hallways, Wakita’s approach learns a compact, environment-agnostic representation, enabling more robust and adaptable navigation systems. His research bridges machine learning and robotics, offering a data-driven solution that enhances the scalability of autonomous systems. Wakita’s work is particularly impactful for researchers developing resource-constrained robots, as it demonstrates how variational autoencoders can replace heuristic feature engineering. By advancing compression techniques for LiDAR data, he contributes to the broader goal of creating more efficient, generalizable, and memory-light autonomous navigation platforms.
Research Focus
Key Achievements
Top Papers
- 1Laser Variational Autoencoder for Map Construction and Self-Localization6 citations · 2018