I will build computer vision solutions, yolo and 3d reconstruction

I build computer vision systems that detect, track, count and reconstruct the real world from images and video.

What I deliver:

Custom object detection trained on your data with YOLO, SAM, PyTorch or TensorFlow

Object tracking and counting for video streams

Image classification with ResNet, MobileNet and Vision Transformers

3D reconstruction from photos and video using NeRF, Gaussian Splatting and COLMAP

Edge and mobile deployment with TensorFlow Lite, LiteRT and ExecuTorch

OCR, pose estimation, segmentation and other vision tasks on request

Clean, deployment ready code with documentation

Why me:

I have shipped vision systems in production, including boat detection and tracking on live video, and an AI kiosk on NVIDIA Jetson running under 200ms response time. I reduced one 3D reconstruction model's GPU memory by 61 percent so it could run on affordable hardware. Gold Medalist Computer Engineer with an IEEE published paper on efficient machine learning.

How it works:

Every project starts with your goal and your data. I will tell you honestly what is achievable before you spend. Message me with sample images or video and I will scope it properly.