Computer Vision Research Engineer Remote, with an optional 1 day a fortnight in London / UK

Remote, with an optional 1 day a fortnight in London.

Are you interested in helping us to craft exceptional experiences for our clients that will have a genuine social good? Are you ready to join a small and experienced team of innovators and make a significant contribution to our success?

About Signapse

Signapse is growing its research team, and is now recruiting for a Computer Vision Research Engineer. This role will expand our video generation pipeline to include Gaussian Splatting and Diffusion for improved realism. Using an approach similar to this paper and this paper, you will train these models on our proprietary sign language data for the task of pose to video generation and video enhancement. This will be using open-source techniques involving data preparation, fine-tuning and inference. Working closely with other Computer Vision Engineers and managed by the Chief Scientist, this role will lead the development of an important part of Signapse’s translation technology. We offer a remote working environment, competitive equity package and opportunity to sculpt an early-stage business building tech for good.

Personal Requirements

• At least 2 years experience with Diffusion models, both inference and training.
• At least 3 years experience using Computer Vision technologies in a commercial environment.
• At least 3 years experience with Python in a commercial environment.
• Experience with deploying CV solutions in a real-world application, with knowledge of both infrastructure and scaling.
• Experience training CV models in a compute cluster e.g. Condor, SLURM.

Example Responsibilities

• Design and build Computer Vision solutions using Gaussian Splatting and Diffusion models for generating Sign Language videos that are realistic and accurate.
• Exploring the landscape of closed- and open-source Diffusion and Gaussian Splatting models.
• Training a Diffusion video generation model on proprietary sign language data for the task of pose to video or video enhancement.
• Training a Gaussian Splatting model on proprietary sign language data for the task of 3D video generation and body movement.
• Improving the scalability and latency of our video generation pipeline.
• Working with the in-house translation team to build the best dataset for our use case.

Knowledge, Skills and Experience

• A Computer Science or other science-based degree background.
• Experience with Gaussian Splatting models based on human appearances e.g. face or upper body.
• Experience with optimising Computer Vision models e.g. fast inference, tensorRt, onnx, distillation, quantisation.
• Research experience such as published academic work or open-source contributions.
• An interest in the latest Computer Vision technology and a willingness to explore the research landscape.
• Understanding of the Deaf community or sign language.
• Git, Docker, AWS knowledge.

24 days’ holiday plus bank holidays, company pension scheme. Competitive compensation and high-value equity packages. Opportunity to work on cutting edge technologies and be involved at the early stages of a high growth business.

Hours

This is a full-time position, with normal virtual office hours of 9am to 6pm, although flexibility is offered to suit reasonable personal circumstances. What matters most to us is strong collaboration, meeting agreed milestones and the delivery of high quality work.

Please note you must have the right to work and live full-time in the UK when you apply for this position.

Equality and Diversity

Signapse is committed to eliminating discrimination and encouraging diversity in our team. Our aim is that our team will be truly representative of all sections of society and each employee feels respected and able to give their best. We’ve created a culture of encouragement and support to enable our employees to focus on what they truly want to achieve for successful career development. Our work-life policies and flexible working practices make our employees feel more in control of their personal and working life.

Any applicants who are native sign language users are guaranteed an interview.

As part of our recruitment process, the Company collects and processes personal information, or personal data, relating to job applicants. We are committed to protecting the privacy and security of your personal information and to meeting our data protection obligations under the General Data Protection Regulation (“GDPR”) and the Data Protection Act 2018. By applying for a position with us you are giving your consent for us to do so.