Computer Vision Research Engineer (PhD Student Preferred)

Computer Vision Research Engineer (PhD Student Preferred)

Remote (India)

Project Overview

Real-time edge AI video analytics system for physical security. Multi-object tracking, temporal behavior analysis, production deployment on consumer hardware.

Target: 10% false positive rate in live deployment.

Tech: YOLOv8 + ByteTrack + TensorRT/DeepStream + Docker + RTSP

Your Role:

Core computer vision research & implementation:

Required Skills

PhD/MS Computer Vision / Machine Learning (PhD students welcome)

YOLOv8 fine-tuning + ByteTrack (GitHub portfolio REQUIRED)

Real-time video processing (OpenCV + PyTorch)

Temporal pattern recognition

TensorRT + DeepStream edge deployment

12-20 FPS on RTX3060/i7 hardware

Bonus: Anomaly detection research, NVIDIA TAO, publications

Deliverables

Multi-object detection + tracking pipeline (90% accuracy)

Temporal validation system (10% FP target)

Production Docker/TensorRT deployment

Research documentation + edge cases

Ideal Profile

Current PhD student (IIT/IIIT/IISc preferred)

200+ hours CV projects

Edge AI deployment experience

English fluency

*Confidential AI Video Analytics Research Project*