Richard Capraru Jun 2026

Dr. Capraru’s research has heavily influenced the ways automotive manufacturers approach multi-sensor fusion. Below is a summary of his most influential contributions:

Served as a visiting doctoral researcher under the elite NTU–TUM–Imperial Global Fellows Programme, collaborating directly with [Professor Emil Constantin Lupu](1.2.3, 1.4.3).

: Standard machine learning models are trained to classify what they see, meaning they struggle to differentiate between a naturally bounced photon and a precisely timed adversarial laser pulse. Weaponizing the Elements: LiDAR Spoofing in Adverse Weather

: Richard’s path has taken him from University College London (UCL) , where he was a Laidlaw Scholar, to major institutions in Singapore, Seoul, Beijing, Hong Kong, and Tokyo. richard capraru

: Currently a PhD candidate at Nanyang Technological University (NTU) and A*STAR in Singapore, his work aims to make self-driving cars safer and more reliable. Story Concept: "The Rain-Reaper"

: Recognizing the data-intensive nature of AI, Capraru developed frameworks for few-shot radar-based recognition

Dr. Capraru’s academic journey is defined by elite global institutions and prestigious fellowships. He began his higher education at , where he graduated with a Bachelor of Engineering (B.Eng.) in Electrical and Electronic Engineering in 2021. During his time at UCL, his early potential was recognized with the competitive Laidlaw Scholarship . : Standard machine learning models are trained to

A primary obstacle for Level 4 and Level 5 autonomous vehicles is navigating dense rain, fog, or snow. While a standard LiDAR sensor maps surrounding topography via infrared light return, rain droplets actively scattering these beams degrade point clouds. Dr. Capraru's pioneering paper, “Rain-Reaper: Unmasking lidar-based detector vulnerabilities in rain” (presented at IROS 2024), exposed how environmental noise acts as a natural camouflage for targeted cyber-attacks.

Search for more specific details on the mentioned.

[Adversarial Laser Emitter] ──> (Low-Power Pulse Hidden in Rain) ──> [Vehicle LiDAR Sensor] │ [Sudden Deceleration / Accident] <── (Perceives Fake Obstacle) <─────────────┘ Enhancing Autonomous Vehicle Defense Frameworks The research demonstrates that common

This public link is valid for 7 days and shares a thread, including any personal information you added. This link or copies made by others cannot be deleted. If you share with third parties, their policies apply. Can’t copy the link right now. Try again later.

: He has explored the "principles of forgetting" in domain-incremental semantic segmentation, particularly for navigating adverse weather conditions. Notable Publications

[Insert information about career, notable achievements, and any relevant accomplishments]

Before shifting fully into autonomous vehicle security, Dr. Capraru vastly expanded the open-source signal processing community's access to clean radar datasets. Alongside co-researchers from UCL and TU Delft, he developed .

The research demonstrates that common, adverse environmental conditions—such as heavy rain, fog, or snow—can actually assist in spoofing attacks, rather than hindering them, by altering the propagation of the light pulses used in the attack.