Computer Vision Scientist

SENIOR COMPUTER VISION SCIENTIST

XYZ Reality are a leading Augmented Reality technology company based in London. We have developed the worlds first AR HMD for Construction that aims to eliminate 2D plans. This allows builders, for the first time, to walk on site and use AR/3D models to construct from start to finish enabling the possibilities to Build it right, First time!

Were looking for an experienced senior computer vision scientist. The AI team at XYZ is concerned with scene understanding in the construction environment to help build impactful client solutions for our Atom and cloud products. The ideal candidate would have worked with neural based computer vision models which aid in scene understanding (such as object detection, semantic segmentation). The candidate would have developed or worked with SOTA vision model architectures and modifying them to perform new downstream domain specific tasks in a commercial product. At XYZ you will work with some of the most talented scientists in the industry and push the boundaries of what computer vision can achieve in complex scene understanding.

This role will report directly into our Computer Vision Lead and will be working with our cutting-edge technology that created our product, the Atom.

Key Responsibilities

• Work with AI team and cross functionally to solve complex vision tasks
• Research discriminative and generative vision SOTA architectures
• Youll be looking to solve the many facets scene understanding in the construction environment
• Making practical R&D decisions in terms of data collection and model training given deadlines
• Willing to learn new things within CV and the construction domain (such as BIM models)

Skills, Knowledge & Expertise

• Msc or Phd in Computer vision or equivalent industry experience.
• Familiar with latest object detection architectures (YOLO family, and other single shot FPN based architectures)
• Familiar with latest semantic and instance segmentation architectures (MaskRCNN, FPN based, DeepLab)
• Experience with neural based generative vision models (GANs, Flow or Diffusion models)
• Experience with large scale model training, and performance evaluations for commercial downstream tasks.
• Python Programming language
• Experienced with a Deep learning development framework, at least one of Pytorch or Tensorflow (Internally we use Pytorch)
• Comfortable in a Linux development environment

Nice to have:

• Experience with traditional computer vision / 2D image processing: e.g. object tracking, trajectory smoothing, morphology, feature tracking.
• Multiview geometry (e.g., 3D reconstruction or SLAM)
• 3D Rendering based data generation for model training
• Experience with Real time 3D graphics tools like Unity, Unreal
• Experience with C++/C# Programming language
• Experience with AI compute accelerators (Nvidia GPU, Intel Movidius or otherwise)
• Model runtime optimisation (quantisation, sparsity)

Benefits

• Competitive Salary
• 25 days paid holidays plus bank holidays
• Company pension
• Flexible working hours
• Private Healthcare
• Easily accessible location (2 min walk from tube station)
• Opportunity to develop skills and learn cutting edge topics along with highly talented team-mates
• You will be heard
• Cycle to work scheme
• Start-up culture, modern, friendly environment including:
• Bi-weekly free lunch
• Fresh fruit every week
• Unlimited tea, coffee, and soft drinks in the office
• Office drinks every Thursday evening
• Summer and Christmas parties
• Casual dress code
• Playing a valuable part in revolutionising the industry.

At XYZ, we believe that our greatest strength lies in the diversity of our team. We are dedicated to creating a workplace culture that embraces and celebrates the unique perspectives, experiences and backgrounds of our employees. Our mission is to empower each member of our team to bring their full, authentic selves to work every day. We believe that diversity drives innovation and creativity, which are essential to our success as a company. By valuing and promoting equality and diversity, we create an environment where all voices are heard, all ideas are considered, and all individuals are empowered to make a meaningful contribution to our shared mission.