ResNet-RS

2000년 도입 · 논문 2편에서 사용

ResNet-RS is a family of ResNet architectures that are 1.7x faster than EfficientNets on TPUs, while achieving similar accuracies on ImageNet. The authors propose two new scaling strategies: (1) scale model depth in regimes where overfitting can occur (width scaling is preferable otherwise); (2) increase image resolution more slowly than previously recommended. Additional improvements include the use of a cosine learning rate schedule, label smoothing, stochastic depth, RandAugment, decreased weight decay, squeeze-and-excitation and the use of the ResNet-D architecture.

출처: Revisiting ResNets: Improved Training and Scaling Strategies

소개 논문: Revisiting ResNets: Improved Training and Scaling Strategies