Bichen Wu, Graduate Student Researcher in the EECS Department at the University of California, Berkeley, presents the “Enabling Automated Design of Computationally Efficient Deep Neural Networks” tutorial at the May 2019 Embedded Vision Summit.
Efficient deep neural networks are increasingly important in the age of AIoT (AI + IoT), in which people hope to deploy intelligent sensors and systems at scale. However, optimizing neural networks to achieve both high accuracy and efficient resource use on different target devices is difficult, since each device has its own idiosyncrasies.
In this talk, Wu introduces differentiable neural architecture search (DNAS), an approach for hardware-aware neural network architecture search. He shows that, using DNAS, the computation cost of the search itself is two orders of magnitude lower than previous approaches, while the models found by DNAS are optimized for target devices and surpass the previous state-of-the-art in efficiency and accuracy. Wu also explains how he used DNAS to find a new family of efficient neural networks called FBNets.
