Computer Vision Engineer

Job Description:

Responsibilities

• Develop computer vision solutions with high-quality, robust, and scalable algorithms and models.
• Collaborate with cross-functional teams to coordinate project activities, manage timelines, and ensure milestones are met.
• Manage the training, evaluation, and fine-tuning of advanced deep learning models for tasks such as object detection, classification, object tracking, and work with models like CLIP and VLM.
• Dockerize applications and work effectively with Git and CI/CD workflows to support smooth development, integration, and deployment.
• Troubleshoot and resolve issues and bugs, while identifying opportunities to improve existing solutions for better scalability and efficiency.

Qualification & Experience

• Experience in Python programming, with proficiency in Git and Docker containerization.
• Experience contributing to the development of computer vision systems for surveillance applications, including violation detection, compliance monitoring, and analytics.
• Familiarity with object detection techniques such as YOLO, DETR, RCNN, and classification algorithms.
• Experience with model conversion tools like ONNX, TensorRT engine, and OpenVINO, including precision optimization and quantization techniques (FP16, INT8).
• Experience with deep learning and vision frameworks such as PyTorch, TensorFlow, OpenCV, and Pillow, building systems optimized for CPU and GPU.
• Ability to implement high-throughput, low-latency inference using tools such as TensorRT, DeepStream, custom CUDA kernels, and Triton Inference Server.
• Good problem-solving skills and eagerness to learn new technologies and techniques.
• Ability to collaborate effectively within a team and communicate technical concepts clearly.
• Good to Have:
• Experience with fine-tuning vision-language models (VLM).
• Knowledge of advanced model quantization techniques beyond FP16 and INT8.
• Familiarity with cloud platforms and deployment tools.
• Understanding of data annotation and augmentation strategies for CV tasks.
• Exposure to real-time video processing and multi-camera systems.