Sandwich Batch Normalization
Code for Sandwich Batch Normalization.
Introduction
We present Sandwich Batch Normalization (SaBN), an extremely easy improvement of Batch Normalization (BN) with only a few lines of code changes.
We demonstrate the prevailing effectiveness of SaBN as a drop-in replacement in four tasks:
- conditional image generation,
- neural architecture search,
- adversarial training,
- arbitrary neural style transfer.
Usage
Check each of them for more information:
Main Results
1. Conditional Image Generation
Using SaBN in conditional generation task enables an immediate performance boost. Evaluation results on CIFAR-10 are shown below:
Model | Inception Score ↑ | FID ↓ |
---|---|---|
AutoGAN | 8.43 | 10.51 |
BigGAN | 8.91 | 8.57 |
SNGAN | 8.76 | 10.18 |
AutoGAN-SaBN (ours) | 8.72 (+0.29) | 9.11 (−1.40) |
BigGAN-SaBN (ours) | 9.01 (+0.10) | 8.03 (−0.54) |
SNGAN-SaBN (ours) | 8.89 (+0.13) | 8.97 (−1.21) |
Visual results on ImageNet (128*128 resolution):
SNGAN | SNGAN-SaBN (ours) |
---|---|
2. Neural Architecture Search
We adopted DARTS as the baseline search algorithm. Results on NAS-Bench-201 are presented below:
Method | CIFAR-100 (top1) | ImageNet (top1) |
---|---|---|
DARTS | 44.05 ± 7.47 | 36.47 ± 7.06 |
DARTS-SaBN (ours) | 71.56 ± 1.39 | 45.85 ± 0.72 |
CIFAR-100 | ImageNet16-120 |
---|---|
3. Adversarial Training
Evaluation results:
Evaluation | BN | AuxBN (clean branch) | SaAuxBN (clean branch) (ours) |
---|---|---|---|
Clean (SA) | 84.84 | 94.47 | 94.62 |
Evaluation | BN | AuxBN (adv branch) | SaAuxBN (adv branch) (ours) |
---|---|---|---|
Clean (SA) | 84.84 | 83.42 | 84.08 |
PGD-10 (RA) | 41.57 | 43.05 | 44.93 |
PGD-20 (RA) | 40.02 | 41.60 | 43.14 |
4. Arbitrary Neural Style Transfer
The model equipped with the proposed SaAdaIN achieves lower style & content loss on both training and testing set.
Training loss:
Training style loss | Training content loss |
---|---|
Validation loss:
Validation style loss | Validation content loss |
---|---|
Acknowledgement
- NAS codebase from NAS-Bench-201.
- NST codebase from AdaIN.