Tunga Aerospace

Computer Vision

One-Hub, Thiruporur · 5 - 7 Years

Posted
3 Apr 2026
Last verified at source
9 hours ago
Apply on Tunga Aerospace
Job Code: CoE-SAI-CV-M-002 Experience: 5 - 7 Years Location: One-Hub, Thiruporur No of Openings: 1 Date of Post: 03-Apr-2026 Applicants: 1 Preferred Experience 5–7 years of experience in computer vision and AI/ML engineering Strong knowledge of deep learning algorithms, real-time systems, and edge deployment workflows Experience with multi-object detection, tracking, and situational awareness systems Exposure to UAV autonomy, EO/IR data processing, and precision landing algorithms Innovation: Exposure to reinforcement learning, SLAM, and sensor fusion for UAV autonomy Compliance Awareness: Familiarity with aerospace/autonomy certification frameworks (DO-178C, DO-254, DGCA UAS guidelines) Leadership: Ability to mentor junior engineers and lead computer vision design reviews Reliability Engineering: Experience in robustness testing, fault tolerance, and redundancy strategies Required Skill Sets & Tools * ML Frameworks: PyTorch, TensorFlow * Optimization Tools: TensorRT, CUDA * Middleware: ROS / ROS 2 * Video/Data Pipelines: GStreamer * Strong debugging, profiling, and documentation skills Qualifications Bachelor’s / Master’s in Computer Science / AI / Robotics Key Responsibility Areas * Deep Learning & Optimization * Develop and optimize deep learning models (quantization, pruning) * Accelerate inference using TensorRT and CUDA * Detection & Tracking * Implement multi-object detection (YOLO-based models) * Develop tracking algorithms (DeepSORT) and classification pipelines * Vision & Navigation Algorithms * Implement visual odometry and precision landing algorithms * Work with EO/IR data for robust perception and situational awareness * Video & Data Pipeline * Optimize real-time video streaming pipelines using GStreamer * Generate synthetic datasets for training and validation * Edge Deployment * Deploy models on edge hardware (Jetson or equivalent) * Perform profiling and optimization for real-time performance