Perception Engineer

Job description

Must Have

• 1+ years of professional experience in C++ and Python.
• Strong hands-on experience with OpenCV and implementation of image

processing algorithms.

• Solid understanding of camera fundamentals, including camera calibration,

intrinsic and extrinsic parameters, focal length, field of view (FOV), lens

distortion, perspective projection, and image formation.

• Good familiarity with video streaming protocols, multimedia pipelines, and

hardware/software encoders (RTSP, H264/H265, GStreamer, FFmpeg, etc.).

Key Responsibilities

• Design, develop, and implement multi-camera real-time streaming systems

on edge devices, supporting features such as object detection, multi-object

tracking, classification, and scene understanding.

• Develop, optimize, and integrate perception pipelines using NVIDIA

DeepStream SDK and GStreamer for high-performance, low-latency video

analytics.

• Optimize computer vision and AI workloads for deployment on NVIDIA Jetson

and other edge computing platforms, considering power, memory, latency, and

computational constraints.

• Integrate USB, CSI, and other cameras into production-grade applications.
• Collaborate closely with hardware, embedded, AI, and software engineering

teams to deliver robust perception solutions.

• Debug and optimize camera interfaces, streaming pipelines, synchronization

issues, and inference performance.

Skills & Requirements

• Bachelor's degree in Computer Science, Electrical Engineering, Robotics,

Electronics, or a related field.

• 1–3 years of professional experience in the Computer Vision domain.
• Strong programming skills in C++ and Python.
• Hands-on experience with NVIDIA DeepStream SDK for building real-time

streaming and video analytics applications.

• Demonstrable experience with GStreamer for multimedia pipeline development

and debugging.

• Extensive experience with OpenCV and classical image processing techniques.
• Experience implementing object detection, multi-object tracking, and image

classification pipelines.

• Good understanding of camera vision fundamentals, and respective

algorithms.

• Familiarity with computer vision algorithms and their practical applications in

real-time systems.

Bonus

• Experience with NVIDIA Jetson platforms (Nano, Xavier, Orin).
• Experience with ROS2
• Experience with CUDA, TensorRT, or ONNX model deployment.
• Familiarity with Docker and edge AI deployment.
• Experience optimizing real-time video processing pipelines for performance and

low latency.