Research Associate - Real World Learning Through Advanced Synthetic Data Approaches

The Alan Turing Institute

The Alan Turing Institute is the UK’s national institute for data science and artificial intelligence whose mission is to make great leaps in data science and artificial intelligence research in order to change the world for the better.

ROLE PURPOSE

To advance the state of the art there is a need to develop AI models that generate a richer representation of the world, much more akin to the human ‘embodied’ experience of learning from continuous streams of data. The challenge is that access to limited (or erroneous) data requires the remainder of the ‘picture’ to be constructed, whilst maintaining the integrity of the world that we are trying to represent.

You will develop novel methods to realise a multi-perspective data environment that enables the fusion of electro-optical and infrared data at multiple spatial and temporal scales. Your work will address three key challenges: synthetic stream data generation, synthetic generation of plausible future scenarios and synthetic generation of edge cases. The techniques developed will help to close the gap between synthetic data generation and real data collections and enable the transferral of AI techniques between different domains. You will pair existing, maturing, and emerging algorithms with synthetic data environments to rapidly evolve autonomous behaviours to detect anomalies and outliers and to perceive meaningful novelty in action. The technical scope of the role includes identifying and advancing synthetic data generation approaches that more richly embody noisy, irregularly sampled, sparse, multi-dimensional geospatial data.

DUTIES AND AREAS OF RESPONSIBILITY

• Design and implement cutting edge AI and mathematical techniques to represent real world environments in the electro-optical and infrared domains, to include multi-sensor fusion approaches for multi-modal streaming data.
• Develop a synthetic data environment for EO/IR data using a gaming engine such as unity, or similar, suitable for interacting with reinforcement learning libraries.
• Compare and improve the performance of AI approaches to the detection and identification of entities in EO and IR data and investigate novel approaches e.g. the application of rough path theory for streaming data.
• Develop AI approaches for the detection of anomalies and outliers in electro-optical and infrared data.
• Deliver robust transparent algorithms that can be reproducibly deployed at scale and develop models that are robust to transferral between domains.
• Publish as lead or co-author in high impact peer reviewed journals and participate in and present at international conferences.

Person Specification

Please see our portal for a full breakdown of the Job Description. 

Terms and Conditions

This full-time post is offered on a 2 year fixed term basis. The annual salary is £40,850-£46,200 (depending on experience) plus excellent benefits: https://www.turing.ac.uk/work-turing/why-work-turing/employee-benefits 

Application procedure

Please see our jobs portal for full details on how to apply. 

Equality Diversity and Inclusion

The Alan Turing Institute is committed to creating an environment where diversity is valued and everyone is treated fairly.  In accordance with the Equality Act, we welcome applications from anyone who meets the specific criteria of the post regardless of age, disability, ethnicity, gender reassignment, marital or civil partnership status, pregnancy and maternity, religion or belief, sex and sexual orientation.  Reasonable adjustments to the interview process can also be made for any candidates with a disability.

Advert information