SEA2Learn · SElf-Adaptive and Automated LEARNing Framework for Smart Sensors
Horizon Europe — Marie Skłodowska-Curie Actions
- Duration
- 2022-09-01 → 2024-08-31
- EU contribution
- €175,920
- Participants
- 2
- Scheme
- HORIZON-TMA-MSCA-PF-EF
Lines connect the coordinator with its partners.
Results in brief
SElf-Adaptive and Automated LEARNing Framework for Smart Sensors
The SEA2Learn research project has investigated lightweight Continual Learning methods for post-deployment adaptation of sensor processing algorithms, i.e. Deep Learning models, embedded inside low-power embedded sensor nodes. The overall aim was to define new methodologies for extreme-edge devices, such as battery-powered smart sensors. These methodologies seek to overcome the current "train-once-deploy-everywhere" paradigm where the local sensor processing algorithms cannot adapt depending on the user and the external environment. The new learning approaches have been examined from both a system and algorithm perspective, particularly in the context of a voice command recognition use case. We focused on tasks with low availability of human-labeled data, which is a common real-world application scenario. To reach the ambitious goal, the SEA2Learn project has pursued the following research Objectives (Obj): Obj1. Design an adaptive smart sensor platform, based on Open Source HW/SW components for real-time adaptation. Obj2. Define efficient mechanisms to continually learn from use-case-specific streams of sensor data under the resource constraints of the tiny devices. Obj3. Demonstrate a fully automated learning process by leveraging unsupervised learning using new data captured with single-sensor and multi-sensor setups.
Data: CORDIS, © European Union
Project objective
Smart Sensors are key components for the upcoming Green and Digital European era. Recently, novel emerging electronics components – such as high energy-efficient many-core application processors featuring a power consumption of few tens of mWs – have enabled high-accurate on-device inference capabilities, i.e. Deep Learning inference, to extract high-level information from sensor data. However, this technology improvement is not sufficient to ensure robust solutions suitable for consumer and industrial applications. The main issue comes from the wide variety in real-world test conditions and, consequently, the lack at design-time of representative (labelled) sensor data, needed to train DL inference networks. For this reason, the currently used “train-once-and-deploy-everywhere” design process for edge intelligence has proved to be weak, even after an endless cyclic procedure involving data collection, model training and in-field testing. This limitation is addressed by the SEA2Learn project by developing energy-efficient and real-time mechanisms to adapt the inference capabilities of resource-constrained smart sensors based on the stimulus from the surrounding environment. The proposed strategy, which is unprecedent in this domain, aims at placing in the same training loop multiple smart sensor nodes that interact with a Learning Agent. The latter will leverage a new class of lightweight methods belonging to the Continual Learning (CL) domain operating on unlabelled multi-sensor data. Thanks to the envisioned SEA2Learn framework, the embedded intelligence can adapt over time based on real-world data, making the design process more robust and 10-100x faster than today. To realize this vision, the fellow’s expertise in HW/SW design for embedded machine learning will be complemented by the Continual Learning knowledge of the hosting research group at KU Leuven and enriched by a tight collaboration with an SME that manufactures IoT platforms for edge computing.
Original text from CORDIS.
Participants
Links
- View on CORDIS
- DOI: 10.3030/101067475
- https://ec.europa.eu/research/participants/documents/downloadPublic?documentIds=080166e5102f2294&appId=PPGMS
- https://ec.europa.eu/research/participants/documents/downloadPublic?documentIds=080166e5f85fe9c5&appId=PPGMS
Data: CORDIS, © European Union