Revisiting ResNets: Improved Training and Scaling Strategies

Novel computer vision architectures monopolize the spotlight, but the impact of the model architecture is often conflated with simultaneous changes to training methodology and scaling strategies. Our work revisits the canonical ResNet (He et al., 2015) and studies these three aspects in an effort to disentangle them. Perhaps surprisingly, we find that training and scaling strategies may matter more than architectural changes, and further, that the resulting ResNets match recent state-of-the-art models. We show that the best performing scaling strategy depends on the training regime and offer 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 (Tan & Le, 2019). Using improved training and scaling strategies, we design a family of ResNet architectures, ResNet-RS, which are 1.7x - 2.7x faster than EfficientNets on TPUs, while achieving similar accuracies on ImageNet. In a large-scale semi-supervised learning setup, ResNet-RS achieves 86.2% top-1 ImageNet accuracy, while being 4.7x faster than EfficientNet NoisyStudent. The training techniques improve transfer performance on a suite of downstream tasks (rivaling state-of-the-art self-supervised algorithms) and extend to video classification on Kinetics-400. We recommend practitioners use these simple revised ResNets as baselines for future research.

PDF Abstract NeurIPS 2021 PDF NeurIPS 2021 Abstract
Task Dataset Model Metric Name Metric Value Global Rank Uses Extra
Training Data
Result Benchmark
Document Image Classification AIP ResNet-RS (ResNet-200 + RS training tricks) Top 1 Accuracy - Verb 83.4 # 1
Image Classification ImageNet ResNet-RS-50 (160 image res) Top 1 Accuracy 84.4% # 299
Number of params 192M # 891
Hardware Burden None # 1
Operations per network pass None # 1
GFLOPs 4.6 # 216
Image Classification ImageNet ResNet-RS-270 (256 image res) Top 1 Accuracy 83.8% # 358
GFLOPs 54 # 432
Semantic Object Interaction Classification Kinetics-700 3D ResNet-50 (RS training) Vid acc@1 78.2 # 1
Image Classification PRImA ResNet-152 2x (RS training) Percentage correct 89.3 # 1

Methods