LA3: Efficient Label-Aware AutoAugment

Mingjun Zhao,Shan Lu,Zixuan Wang,Xiaoli Wang,Di Niu

DOI: https://doi.org/10.1007/978-3-031-19803-8_16

2023-04-20

Abstract:Automated augmentation is an emerging and effective technique to search for data augmentation policies to improve generalizability of deep neural network training. Most existing work focuses on constructing a unified policy applicable to all data samples in a given dataset, without considering sample or class variations. In this paper, we propose a novel two-stage data augmentation algorithm, named Label-Aware AutoAugment (LA3), which takes advantage of the label information, and learns augmentation policies separately for samples of different labels. LA3 consists of two learning stages, where in the first stage, individual augmentation methods are evaluated and ranked for each label via Bayesian Optimization aided by a neural predictor, which allows us to identify effective augmentation techniques for each label under a low search cost. And in the second stage, a composite augmentation policy is constructed out of a selection of effective as well as complementary augmentations, which produces significant performance boost and can be easily deployed in typical model training. Extensive experiments demonstrate that LA3 achieves excellent performance matching or surpassing existing methods on CIFAR-10 and CIFAR-100, and achieves a new state-of-the-art ImageNet accuracy of 79.97% on ResNet-50 among auto-augmentation methods, while maintaining a low computational cost.

Computer Vision and Pattern Recognition,Artificial Intelligence

What problem does this paper attempt to address?

### Problems Addressed by the Paper

The paper aims to address the optimization of data augmentation strategies in the training of deep neural networks. Specifically:

1. **Limitations of Existing Methods**:
- Most existing automated data augmentation methods focus on constructing a unified strategy applicable to the entire dataset without considering the differences in samples or categories, which may lead to suboptimal solutions.

2. **Proposed New Method**:
- A new algorithm named "Label-Aware AutoAugment (LA3)" is proposed, which utilizes label information to learn independent augmentation strategies for data samples with different labels.
- LA3 consists of two stages: the first stage evaluates effective augmentation operations for each label using a Bayesian optimization-assisted neural predictor; the second stage constructs a composite augmentation strategy based on the principle of minimum redundancy and maximum reward.

3. **Experimental Results**:
- LA3 outperforms existing static and dynamic automated augmentation methods on the CIFAR-10, CIFAR-100, and ImageNet datasets, and also shows significant advantages in computational efficiency.

In summary, the paper proposes the LA3 algorithm to overcome the limitations of existing data augmentation strategies in considering sample differences, thereby improving model performance and generalization ability.