Computer Vision Engineer

About the Role

Role: Computer Vision Engineer

Employment Type: Full-Time

Key Responsibilities

• Develop computer vision solutions for detection, classification, segmentation, tracking, and related perception tasks.
• Prepare and manage datasets, including data collection, annotation, preprocessing, and augmentation.
• Train, fine-tune, validate, and benchmark deep learning models.
• Analyze model performance, failure cases, false positives/negatives, and dataset limitations.
• Optimize models and algorithms for real-time inference and deployment.
• Develop and integrate performance-critical components using C++.
• Work with GPU-accelerated training and inference using NVIDIA/CUDA-based platforms where applicable.
• Maintain clear documentation of experiments, results, and model performance.

Requirements & Tech Stack

• Strong fundamentals in computer vision, image processing, and deep learning.
• Understanding of CNNs and common vision tasks such as object detection, classification, segmentation, and tracking.
• Hands-on experience with OpenCV.
• Good programming skills in C++ and Python.
• Familiarity with PyTorch or TensorFlow.
• Understanding of model training, validation, testing, and evaluation metrics.
• Basic understanding of GPU-based training and inference.
• Comfortable working in Linux and Git-based development environments.
• Strong analytical, debugging, and problem-solving skills.

Qualifications

• M.Tech/ Ph.D

Good to Have

• Familiarity with ROS2.
• Exposure to sensors such as cameras, LiDAR, radar, IMU, or GNSS.
• Understanding of camera calibration and multi-sensor calibration.
• Exposure to CUDA, TensorRT, NVIDIA Jetson, or DeepStream.
• Experience with real-time perception or robotics applications.