Computer Vision - Perception Engineer

Computer Vision / Perception Engineer

Robotics, 3D Vision & Edge AI
For one of our technology clients, we are looking for a Computer Vision / Perception Engineer to work on advanced AI systems operating in the physical world. This is not a role focused only on training models in notebooks or improving offline benchmarks. The project is about building perception systems that help machines understand real environments: objects, movement, depth, space, geometry, uncertainty and context. You will work at the intersection of Computer Vision, Deep Learning, Robotics, 3D perception and Edge AI, contributing to systems that can be deployed outside the lab, integrated with real hardware and tested against real-world constraints such as latency, lighting, occlusions, sensor noise and limited compute.

About the project
Our client is developing next-generation AI solutions for systems that need to perceive, interpret and act in complex physical environments.
The team is working on technologies that enable machines to “see” and understand the world around them. This includes object detection, segmentation, pose estimation, tracking, depth perception, point cloud processing, model optimization and deployment of AI models on edge devices.
The project combines the research-driven side of AI with practical engineering. You will not only build models, but also help bring them closer to production: optimize inference, improve robustness, validate results on real data and collaborate with software, robotics and embedded teams.
This is a great opportunity for someone who wants to work on AI beyond the screen: systems connected to cameras, sensors, machines, robots, industrial environments or autonomous platforms.

What you will work on
You will be involved in:
• Designing and developing Computer Vision algorithms for real-world applications
• Building and improving Deep Learning models for detection, segmentation, classification, tracking or pose estimation
• Working with 2D images, depth cameras, 3D data, point clouds or multi-sensor inputs
• Improving model performance, robustness and inference latency
• Validating models on real-world datasets and edge cases
• Collaborating with robotics, software and embedded teams to integrate AI models into broader systems
• Supporting deployment of models on edge devices, GPUs or embedded platforms
• Optimizing models using tools such as ONNX, TensorRT, CUDA, OpenVINO or similar technologies
• Helping transform prototypes into reliable systems that can operate outside controlled environments

What we are looking for
We are looking for someone with:
• Strong experience with Python
• Practical experience in Computer Vision and Deep Learning
• Experience with PyTorch or TensorFlow
• Good understanding of image processing and modern vision architectures
• Experience with models for object detection, segmentation, classification, tracking or pose estimation
• Hands-on experience with tools such as OpenCV, NumPy, scikit-image or similar libraries
• Ability to work with real datasets and evaluate model quality beyond simple benchmark metrics
• Good understanding of geometry, camera calibration, transformations or 3D perception basics
• Solid engineering mindset and ability to turn AI prototypes into working solutions
• English allowing you to work in an international technical environment

Nice to have
It would be great if you also have experience with:
• Robotics, autonomous systems or industrial automation
•

C++

•

ROS / ROS2

• 3D Computer Vision, point clouds, PCL, Open3D
• LiDAR, RGB-D cameras, stereo cameras or multi-camera systems
• Pose estimation, depth estimation, scene reconstruction or visual tracking
• Edge AI deployment on NVIDIA Jetson or similar platforms
•

CUDA, TensorRT, ONNX, OpenVINO

• Model quantization, pruning or inference optimization
• Docker, Linux and production-oriented ML workflows
• Simulation environments such as Isaac Sim, Gazebo, Unity or Unreal Engine

Tech stack

Core:
Python, Computer Vision, Deep Learning, PyTorch/TensorFlow, OpenCV, NumPyNice to have:
C++, ROS2, CUDA, TensorRT, ONNX, NVIDIA Jetson, Open3D, PCL, Docker, Linux

Why this role is interesting
This role gives you the opportunity to work on AI systems that are much closer to the physical world than typical ML projects.
You will work on problems where accuracy is only one part of the challenge. The system also needs to be fast, stable, robust and useful in changing real-world conditions. You will deal with real cameras, imperfect data, physical constraints and deployment requirements.
If you enjoy combining Computer Vision, Deep Learning and practical engineering, this role offers a strong technical challenge and a chance to contribute to one of the most exciting areas of modern technology: Physical AI.