Master's Thesis: On-device Learning for resource-constrained IoT devices

We have an open position for a dedicated master student to join us at the Connected Intelligence research unit, within the Computer Science department at RISE.

Background and Purpose
Machine learning (ML) has catalyzed the design and development of various intelligent systems over wide application domains. Many efforts have been made to deploy ML models on resource-constrained devices as well. More recently, the concept of on-device learning (ODL) has gained attention, which means performing both inference and training on local devices. Local devices mostly refer to resource-constrained devices, ranging from smartphones to low-power IoT devices. For example, low-power IoT devices are normally only equipped with memory in the kilobyte range, which brings challenges to the feasibility of ODL. Meanwhile, ODL also comes with the following two main benefits: firstly, ODL alleviates data drift via online adaptation for different deployment scenarios. Secondly, ODL preserves privacy, saves bandwidth, and reduces latency and energy since there is no need to upload data to the cloud and/or download an updated model back.

Thesis Description
The thesis will focus on investigating the performance, power consumption, memory usage, and computing resources usage of the ODL system on resource-constraint IoT devices. After a thorough literature review of related research works, an ODL system will be designed, implemented, and evaluated.
Terms:
Start time: as soon as possible.
Scope: 30hp (20 weeks)
Location: RISE Computer Science, Stockholm (Kista). Options to partially work remotely.

Who are you?
We expect you to have good programming skills with C/C++, experience in applying machine learning models, and an interest/curiosity in IoT and embedded systems.

Welcome with your application!
To know more, please contact Prof. Thiemo Voigt (thiemo.voigt@ri.se), and Shuai Zhu (shuai.zhu@ri.se). Applications should include a brief personal letter, CV, recent grades, and a code example. Candidates are encouraged to send in their application as soon as possible but at the latest by the 15th of October 2024. Suitable applicants will be interviewed as soon as applications are received.

Keywords: Master thesis, Machine Learning, On-device learning, IoT, RISE, Stockholm