SONATA: Self-adaptive Evolutionary Framework for Hardware-aware Neural Architecture Search
Halima Bouzidi, Smail Niar, Hamza Ouarnoughi, El-Ghazali Talbi

TL;DR
SONATA introduces a self-adaptive evolutionary framework for hardware-aware neural architecture search, leveraging learned importance of design parameters and surrogate models to improve search efficiency and optimize neural network performance on IoT devices.
Contribution
It proposes a novel self-adaptive evolutionary algorithm that uses surrogate models and reinforcement learning to guide neural architecture search for IoT hardware.
Findings
Achieved up to 0.25% accuracy improvement on ImageNet-1k.
Gained up to 2.42x reductions in latency and energy consumption.
Outperformed NSGA-II with 93.6% Pareto dominance.
Abstract
Recent advancements in Artificial Intelligence (AI), driven by Neural Networks (NN), demand innovative neural architecture designs, particularly within the constrained environments of Internet of Things (IoT) systems, to balance performance and efficiency. HW-aware Neural Architecture Search (HW-aware NAS) emerges as an attractive strategy to automate the design of NN using multi-objective optimization approaches, such as evolutionary algorithms. However, the intricate relationship between NN design parameters and HW-aware NAS optimization objectives remains an underexplored research area, overlooking opportunities to effectively leverage this knowledge to guide the search process accordingly. Furthermore, the large amount of evaluation data produced during the search holds untapped potential for refining the optimization strategy and improving the approximation of the Pareto front.…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsEvolutionary Algorithms and Applications · Neural Networks and Applications · Machine Learning and Data Classification
