Multi-objective Reinforced Evolution in Mobile Neural Architecture Search
| Autor: | Xu Ruijun, Bo Zhang, Xiangxiang Chu |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Computer science
business.industry Evolutionary algorithm 02 engineering and technology FLOPS 01 natural sciences 010305 fluids & plasmas Domain (software engineering) Task (computing) 0103 physical sciences Genetic algorithm 0202 electrical engineering electronic engineering information engineering Code (cryptography) Reinforcement learning 020201 artificial intelligence & image processing Artificial intelligence business Mobile device |
| Zdroj: | Computer Vision – ECCV 2020 Workshops ISBN: 9783030668228 ECCV Workshops (4) |
| DOI: | 10.1007/978-3-030-66823-5_6 |
| Popis: | Fabricating neural models for a wide range of mobile devices is a challenging task due to highly constrained resources. Recent trends favor neural architecture search involving evolutionary algorithms (EA) and reinforcement learning (RL), however, they are separately used. In this paper, we present a novel multi-objective algorithm called MoreMNAS (Multi-Objective Reinforced Evolution in Mobile Neural Architecture Search) by leveraging good virtues from both sides. Particularly, we devise a variant of multi-objective genetic algorithm NSGA-II, where mutations are performed either by reinforcement learning or a natural mutating process. It maintains a delicate balance between exploration and exploitation. Not only does it prevent the search degradation, but it also makes use of the learned knowledge. Our experiments conducted in Super-resolution domain (SR) deliver rivaling models compared with some state-of-the-art methods at fewer FLOPS (Evaluation code can be found at https://github.com/xiaomi-automl/MoreMNAS). |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|