Computer Vision Engineer

Key Responsibilities:

• Develop and implement computer vision pipelines for object detection, tracking, and marker detection.
• Utilize PyTorch and frameworks like YOLO, Ultralytics, MMDetection, SuperGradients for model training and deployment.
• Apply multi-frame temporal reasoning and feature embedding for identity tracking and matching.
• Optimize GPU performance using NVIDIA hardware, including TensorRT, CUDA, and batching strategies.
• Scale inference workloads using distributed frameworks such as Ray or Dask.
• Implement ReID models, cosine similarity scoring, and association logic for accurate tracking.
• Handle real-time video ingestion and streaming using RTSP, Kafka/MQTT, and asynchronous queues.
• Troubleshoot detection issues related to occlusion, blur, and orientation by tuning confidence thresholds.
• Deploy CV services in Docker containers, preferably with Ray Serve or FastAPI-based microservices.
• Store and validate event data using relational databases like PostgreSQL/MySQL