I'm Hendrik,
I currently work as a data analyst in the operations team at Mainspring Energy Inc. to predict fleet reliability using data science.
Prior to that, I received two masters degrees in Mechanical Engineering from UC Berkeley and Arts et MĂ©tiers ParisTech (France).
I'm interested in robotics and deep learning and advise robotics projects at UC Berkeley on the transfer of skills from humans to robots.
During my free time, I enjoy surfing unridden waves between SF and Santa Cruz.
Teleoperation of low-cost robotic manipulators remains challenging due to the complexity of mapping human hand articulations to robot joint commands. We present a system that enables robot teleoperation from a single egocentric RGB-D camera mounted on 3D-printed glasses. The pipeline detects 21 hand landmarks per hand using MediaPipe Hands, deprojects them into 3D via depth sensing, transforms them into the robot coordinate frame, and solves a damped-least-squares inverse kinematics problem in PyBullet to produce joint commands for the 6-DOF SO-ARM101 robot....
We compare two U-Net architectures — a full 3D convolutional U-Net and a 2.5D slice-conditioned U-Net — for brain MRI super-resolution using the Elucidated Diffusion Model (EDM) framework. Trained on just 59 subjects from the FOMO60K dataset, the 3D model achieves 37.77 dB PSNR and 0.996 SSIM on 2x super-resolution, surpassing pretrained EDSR and Swin2SR baselines by over 2 dB. Code: GitHub | Weights: HuggingFace | Dataset: FOMO60K / NKI Demo: Low Resolution to High Resolution MRI Methods We adopt the EDM framework (Karras et al....
This project improves object detection in autonomous vehicles by integrating radar and camera data. The pipeline processes and aligns data from both sensors, detects objects using YOLO, and clusters radar points with DBSCAN. The merged results offer a precise and reliable view of the vehicle’s surroundings, enhancing detection accuracy and safety in complex environments. Code : https://github.com/chichonnade/ZendarComputerVisionCapstone Demo Driving scene showing obstacles detected by our pipeline Raw data from radar and camera sensors Motivation Data Fusion Validation:...
Author: Hendrik Chiche Supervisor: Eric Chevalier Background and Motivation Every airborne vehicle is subject to continuous friction with atmospheric particles and air molecules, leading to triboelectric charge separation and accumulation on the fuselage. This effect can be modeled analogously to a capacitor: $$ V = \frac{Q}{C} $$where: \( V \) is the potential difference generated, \( Q \) is the accumulated charge, \( C \) is the effective capacitance of the airframe....