ResearchBib Share Your Research, Maximize Your Social Impacts
Sign for Notice Everyday Sign up >> Login

Postdoctoral position in machine learning and activity recognition, University of Sussex

Country : UK - United Kingdom

Website :


The Wearable Technologies Lab at the University of Sussex is looking for a postdoctoral researcher to work with Dr Daniel Roggen on the research-intensive project "Mobile Activity Recognition with Machine Learning and Deep Learning".
The Wearable Technologies Lab is in the process of collecting the world's largest dataset of modes of transportation of mobile phones users in the wild, with accurate annotations.
This highly multimodal dataset may surpass 1TB once completed in the summer. It sets a new standardfor the community to benchmark activity recognition methods and will be released publicly. Reference publications using this dataset have a strong potential for high citations.
** Role **
Your role will be to analyse the dataset and to develop reference activity recognition methods based on classical machine learning to recognise modes of transport and user activities. A particular research focus is to uncover computationally efficient approaches which use single or multiple sensor modalities, possibly exploiting dynamic tradeoffs between features and modalities depending on energy availability and performance goals.
In a second phase you will investigate novel approaches based on deep learning to exploit the temporal dynamics of the dataset and learn suitable representations from the data which may further enhance performance. You will analyse learned features, identifying most relevant subsets to minimise computational costs, and finding ways to efficiently exploit the temporal dynamics within the dataset.
This project is based at the growing Wearable Computing Group at the University of Sussex ( This project gives the opportunity to collaborate with a large multinational interested in behaviour analytics which is associated with this project. It also gives possibility to patent some of the research outcomes.
** Key Requirements **
This post is well suited to a highly motivated individual with excellent technical skills and with a willingness to operate in a dynamic research environment within an international team.
Candidates should have a PhD (or will shortly be assessed for a PhD) in Computer Science, Mathematics or Electrical Engineering with a strong background in (one or more of) Machine Learning, Time Series Analysis, Signal Processing, Data Mining and ideally Deep Learning.
An established expertise in Activity Recognition and Wearable/Mobile Computing is desired. The candidate should have a strong interest in the combination of theoretical and experimental research.
** Background **
The Wearable Technologies Lab of Dr. Daniel Roggen at the University of Sussex develops novel wearable sensors and methods to recognize and understand human activities with applications to sports, healthcare, entertainment and industry.
Our vision are wearable systems capable of lifelong learning and adaptation to their users. Research outcomes tie in to artificial intelligence and more broadly to action perception.
** Advantages and career development **
This position is ideally suited for somebody who wants to broaden his/her knowledge in wearable or ubiquitous computing and activity recognition and that is interested in deep learning research.
The project is suitable for high impact publications as they will lead to reference performance benchmarks on the world's largest dataset of modes of transportation and will allow the candidate to gain international visibility. The project gives a chance to collaborate with a major industrial player interested in behavioural analytics.
The candidate will be supported in applying for grants to support his/her further career development.
** More information **
For further information and informal inquiries contact Dr Daniel Roggen:
More details about the nature of the work can be inferred from past publications:
More information about the Wearable Technologies Lab:
Applications should be accompanied by a full CV and a statement of how you envisage your role.
Apply online here:

Last modified: 2017-05-18 23:00:45