Machine Learning Engineer – Computer Vision

Job Title: Computer Vision Engineer

Experience: 3-6 years

Location: Gurgaon

Job Type: Full-time

Key Responsibilities

• Design and implement real-time, accurate 2D object detection on BGR and grayscale video streams.
• Develop 3D object detection algorithms using LiDAR point-cloud data.
• Perform depth mapping, SLAM, and point-cloud analysis for spatial understanding.
• Lead image/video labelling efforts and build training datasets for models like YOLOv5/v7/v11.
• Convert research prototypes into optimized, production-ready code for deployment.
• Work closely with the UI/Frontend and integration teams.
• Develop robust Python scripts and C++ modules for model training, testing, and deployment.
• Conduct sensor calibration including camera-LiDAR alignment and transformation into real-world frames.

Required Skills & Qualifications

• Basic Knowledge in computer vision, image processing, and machine learning.
• Proficiency in Python and C++ with a strong foundation in DSA, multi-threading, multi-processing, and IPC.
• Deep understanding of OpenCV, PyTorch, NumPy, and related libraries.
• Experience working with depth cameras, 3D LiDAR, and point cloud data.
• Comfortable developing on Linux platforms and deploying on embedded systems (e.g., NVIDIA Jetson).
• Familiarity with model training, optimization, and deployment pipelines.
• Strong debugging and performance optimization skills, especially on resource-constrained devices.
• Hands-on experience in Linux & Embedded Linux (Must)
• Understanding of scene and camera projection matrices.

Technologies & Tools Used

• Sensors: 3D LiDAR, HD Cameras, Thermal Cameras, Depth Cameras (e.g., Intel RealSense, ZED)
• Platforms: NVIDIA Jetson Nano/Xavier, Orin, Raspberry Pi, Ubuntu
• Frameworks: YOLOv5/v7/v11, PyTorch, OpenCV, Open3D, ROS/ROS2
• Languages: Python, C++, Bash
• Others: Git, Docker, Jetpack, Linux, SLAM toolkits