< Back to previous page

Project

Advances in Neural Architecture Search

Currently, a dominating paradigm in artificial intelligence is Deep (Convolutional) Neural Networks. The development of deep learning has broken benchmark results across a wide range of domains starting with the ImageNet Challenge in 2012, and currently defines the state-of-the-art in application areas such as computer vision, speech recognition, natural language processing, etc. The current state of the art in machine learning involves either manual tuning of network architectures, or intensive neural architecture search methods that can have immense computational and energy demands. In this research, we will investigate methods to perform neural architecture search with improved efficiency through intelligent optimization strategies.
Date:10 Apr 2020 →  Today
Keywords:deep learning, neural architecture search
Disciplines:Computer vision
Project type:PhD project