Senior Data Scientist (Computer Vision)

A genuinely interesting opportunity has come up to join a rapidly scaling start-up working on advanced computer vision, machine learning and high-performance video systems.

They're looking for a Senior Data Scientist specialising in Computer Vision & Object-Tracking to join as a foundational hire. You'll take ownership of applied research and modelling while helping shape the technical direction and wider team as the capability grows over the next 12–18 months. 

This isn't traditional Data Science. You'll solve complex, real-world video problems across detection, tracking, pose estimation, temporal modelling and movement analysis, with the freedom to challenge existing approaches and develop better methods where standard models aren't good enough.

What You'll Do

• Own applied Computer Vision research and modelling across a growing suite of products.
• Develop and evaluate models across object detection, multi-object tracking, ReID, pose estimation and video understanding.
• Develop 2D and 3D pose estimation approaches for human movement analysis.
• Explore temporal and trajectory-based approaches to understand actions and movement across video.
• Turn ambiguous real-world problems into measurable hypotheses, experiments and solutions.
• Own dataset design, annotation strategy and rigorous model evaluation.
• Identify failure modes in existing approaches and develop or adapt methods to improve performance.
• Prototype primarily in Python and PyTorch, working closely with Software Engineers to take successful research into production.
• Help establish the research practices and technical foundations for a growing Computer Vision capability.

What You'll Bring

• PhD in Computer Vision, Machine Learning, AI, Computer Science, Engineering, Applied Mathematics, Physics or a closely related quantitative field.
• A genuine scientific mindset – hypothesis-driven, experimental and evidence-led in your approach.
• Strong hands-on experience developing Computer Vision or video ML models using Python and PyTorch.
• Research depth across several areas such as detection, multi-object tracking, ReID, pose estimation, action recognition, temporal modelling or geometric Computer Vision.
• Strong experimental design, dataset design and model evaluation skills.
• Comfortable working with messy, real-world video and problems where there isn't an obvious off-the-shelf solution.
• Strong quantitative reasoning and the ability to clearly communicate research outcomes and technical trade-offs.

Highly regarded

• Multi-object tracking and person ReID.
• 3D pose estimation, multi-view geometry or 3D reconstruction.
• Temporal/sequence modelling and video transformers.
• Trajectory modelling, probabilistic or physics-informed approaches.
• Biomechanics or human movement experience.
• Camera geometry and calibration.
• Experience with large-scale annotation or dataset curation.
• Experience translating academic research into production applications.

If you have any questions about this role or others we have available you can email Jack directly via [email protected]

For The Latest Jobs, Tech News or to 'Introduce A Friend' for a $1000 Referral Fee - Follow Us On Linkedin:

https://www.linkedin.com/company/thedrivegroup 

TheDriveGroup is 100% committed to improving meaningful diversity in the technology industry.

We partner with clients who embrace diversity and seek candidates across all backgrounds, abilities, genders, sexualities, cultures, and faiths.

If you have the right skills and attitude we would love to hear from you!