Job highlights

- Solid understanding of real-time image and point-cloud processing,with evaluation using mAP,IoU,FPS,and latency metrics. Hands-on experience with ROS 2 (Humble preferred). LiDAR processing,radar-based object tracking and multi-modal sensor fusion. Solid programming skills in Python and modern C++ (C++17+)

- Required Skills

Job Summary

We are looking for a skilled Perception AI Engineer to develop, optimize, and deploy real-time object detection and terrain perception models on our ADAS-ACU platform used in autonomous tactical ground vehicles. You will be responsible for delivering a reliable multi-sensor perception pipeline using EO/IR cameras, LiDAR, Radar with real-time performance constraints (<100 ms latency). This role directly enables situational awareness and safety features for the autonomy stack.

You will collaborate closely with embedded, fusion, and planning engineers to translate sensor data into mission-ready intelligence.

Responsibilities

- Develop and optimize deep learning models (YOLOv5/v8, DeepLabV3+ or similar) for object detection and terrain classification

- Perform model optimization (quantization, pruning, compression) for real-time edge inference on ADAS controllers

- Deploy and optimize perception pipelines on embedded AI platforms (NXP eIQTM, Hailo-8, CPU/NPU balancing)

- Integrate camera, LiDAR, radar, and IMU data with calibration, synchronization,



and point-cloud processing (PCL/Open3D)

- Build and maintain ROS 2 perception nodes with QoS tuning and standardized messaging for planner integration

- Validate performance using simulation, rosbag replay, and on-vehicle field testing under real-time constraints

Required Skills

- 5-8 years of hands-on experience in Computer Vision / Perception for real-time systems

- Strong expertise in deep learning frameworks and deployment pipelines (PyTorch, TensorRT, ONNX)

- Proven experience in model optimization techniques including 8-bit quantization, pruning, and layer fusion

- Solid understanding of real-time image and point-cloud processing, with evaluation using mAP, IoU, FPS, and latency metrics

- Hands-on experience with ROS 2 (Humble preferred)

- LiDAR processing, radar-based object tracking and multi-modal sensor fusion

- Strong programming skills in Python and modern C++ (C++17+); CUDA, CVAT/SuperAnnotate, RViz, rosbag, GStreamer, and profiling tools are a plus

Educational Qualifications

- Bachelor s or Master s degree in Computer Science, Electronics, Embedded Systems, or related fields

Disclaimer: This job posting has been aggregated from external source. Role details, content, and availability are subject to change. Applicants are advised to confirm the latest information directly on the company website before applying.

📌 Perception & AI Engineer - Autonomous UGV Systems (Bengaluru)
🏢 OPTM MEDIA SOLUTIONS PRIVATE
📍 Bengaluru