Summary
Mecka AI builds data infrastructure for robotics and embodied AI. The Research Scientist will design and train models for 3D human pose estimation, body mesh recovery, kinematic tracking, and human-scene interaction, using proprietary motion data. The role also involves scaling training across multi-GPU clusters and developing models for occlusion, complex motion, and physical contact. Applicants should have deep learning and 3D computer vision expertise, experience training large vision models from scratch, and strong knowledge of PyTorch and human body models.
Sign up free to read the full posting.
The employer's posting, the pay, the skills and the way to apply are all there. It takes a minute, and it stays free.