Computer Vision Engineer for Offline 3D Body Reconstruction
We are seeking an experienced Computer Vision / 3D Reconstruction Engineer to develop a standalone, offline software for accurate, contactless human body measurements tailored for bespoke tailoring.The software will process multi-view photos or slow-motion videos to reconstruct a metric-accurate 3D human mesh and automatically compute standard tailoring measurements such as chest, waist, hips, inseam, shoulder width, sleeve length, and neck circumference.Remote candidates across India are welcome, with preference for developers near Navi Mumbai, Mumbai, or Pune for occasional in-person calibration and testing.Key Responsibilities:Build an offline pipeline for 3D body geometry reconstruction from images or video frames.Implement parametric human body estimation models (e.g., SMPL, SMPL-X).Establish metric scaling and calibration protocols targeting ±2 mm accuracy.Automate extraction of circumferential and point-to-point body measurements from 3D meshes.Deliver a lightweight, standalone offline desktop application for Windows or Linux.Technical Requirements:Expertise in 3D Computer Vision, Human Mesh Recovery, Photogrammetry, and Surface Geometry.Proficiency with Python/C++, OpenCV, Open3D, PyTorch3D, Trimesh, or MediaPipe.Strong domain knowledge of parametric body models and 3D mesh processing techniques.Deliverables for Phase 1 (PoC):Working pipeline generating closed 3D body mesh from sample inputs.Automated extraction of 8–10 key tailoring dimensions with error analysis.Clean, well-documented source code and basic offline GUI.To apply, please share relevant past work, your approach to achieving ±2 mm accuracy, willingness for occasional onsite visits, and an estimated timeline with milestone-based quote.