Designing Network Design Spaces

In this work, we present a new network design paradigm. Our goal is to help advance the understanding of network design and discover design principles that generalize across settings... (read more)

PDF Abstract CVPR 2020 PDF CVPR 2020 Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Image Classification ImageNet RegNetY-4.0GF Top 1 Accuracy 79.4% # 77
Number of params 20.6M # 47
Image Classification ImageNet RegNetY-1.6GF Top 1 Accuracy 78.8% # 89
Number of params 11.2M # 53
Image Classification ImageNet RegNetY-800MF Top 1 Accuracy 76.3% # 120
Number of params 6.3M # 63
Image Classification ImageNet RegNetY-600MF Top 1 Accuracy 75.5% # 128
Number of params 6.1M # 64
Image Classification ImageNet RegNetY-8.0GF Top 1 Accuracy 79.9% # 73
Number of params 39.2M # 35

Methods used in the Paper