Discovering Robust Convolutional Architecture at Targeted Capacity: A Multi-Shot Approach
Considering scenarios with capacity budget, we aim to discover adversarially robust architecture at targeted capacities. However, one-shot NAS methods favor topologies with larger capacities in the supernet. And the discovered topology might be suboptimal when augmented to the targeted capacity. We propose a novel multi-shot NAS method to address this issue and explicitly search for robust architectures at targeted capacities.
Dec 22, 2020
