: During his doctoral program, he worked with the Institute for Infocomm Research (I
What makes a subject worthy of a long-form article is not just his track record, but his methodology. In an era of "hustle culture" and get-rich-quick schemes, Capraru represents the counter-movement: slow, deliberate, structural growth. He is the strategist you call when you are tired of putting out fires and want to build a fireproof building.
: A significant portion of his work explores how rain and other weather conditions affect LiDAR and radar detectors. He has developed approaches to "unmask" vulnerabilities and overcome "catastrophic forgetting" in object detection models during inclement weather. Security and Spoofing richard capraru
Published via IEEE RadarConf25, offering deep transfer-learning solutions to keep multi-sensor systems stable after retraining. Future Horizon: Neurointelligence and Embodied AI
Dr. Richard Capraru is a researcher currently affiliated with the at the University of Tokyo . His expertise focuses on the intersection of robust autonomous systems, adversarial perception, predictive models, and AI agency. Professional Profile : During his doctoral program, he worked with
While LiDAR is known to be relatively robust to environmental interference, studies suggest that intensity and the number of detected points can be attenuated by rain. Capraru's research, such as "Leveraging Adverse Weather for Enhanced LiDAR Spoofing in Autonomous Driving," takes this further by exploring how these atmospheric impacts can be intentionally manipulated, rather than just observed as technical limitations. Key Collaborators
Pioneering Research: LiDAR, Weather, and Adversarial Vulnerabilities : A significant portion of his work explores
It is possible that:
: Earlier in his career, he contributed to the development of
Curious about the datasets behind these breakthroughs? Check out the latest on Dop-NET to see how we're benchmarking the next generation of radar micro-Doppler signatures.
: Injecting realistic physical weather distortions into clear point clouds to broaden the model’s exposure.