Adversarial Auto-Augment with Label Preservation: A Representation Learning Principle Guided Approach

Authors: Kaiwen Yang, Yanchao Sun, Jiahao Su, Fengxiang He, Xinmei Tian, Furong Huang, Tianyi Zhou, Dacheng Tao

NeurIPS 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In this section, we apply LP-A3 as a data augmentation method to several popular methods for three different learning tasks, i.e., (1) semi-supervised classification; (2) noisy-label learning and (3) medical image classification. In all the experiments, LP-A3 can (1) consistently improve the convergence and test accuracy of existing methods and (2) autonomously produce augmentations that bring non-trivial improvement even without any domain knowledge available. ... Semi-supervised learning To evaluate how LP-A3 improves the learning without sufficient labeled data, we conduct experiments on semi-supervised classification on standard benchmarks including CIFAR [22] and STL-10 [10] where only a very small amount of labels are revealed. ... We report the results in Table 3, where Rand Augment designed for natural images fails to improve the performance in this scenario. In contrast, LP-A3 does not rely on any domain knowledge brings improvement to all the datasets, especially for Oct MNIST where the improvement is over 1%.
Researcher Affiliation Collaboration University of Science and Technology of China1; University of Maryland, College Park2 Institute of Artificial Intelligence, Hefei Comprehensive National Science Center3 JD Explore Academy4; The University of Sydney5
Pseudocode Yes The pseudo-code of plugging LP-A3 into the representation learning procedure with TCS-based data selection is provided in Algorithm 1.
Open Source Code Yes Code is available at: https://github.com/kai-wen-yang/LPA3.
Open Datasets Yes We conduct experiments on semi-supervised classification on standard benchmarks including CIFAR [22] and STL-10 [10]... medical image classification tasks from Med MNIST [50]
Dataset Splits Yes More experimental details can be found in the Appendix. ... Did you specify all the training details (e.g., data splits, hyperparameters, how they were chosen)? [Yes] See Aappendix C
Hardware Specification No The paper states in its checklist that it did not include the total amount of compute and the type of resources used. While it mentions 'GPU cluster built by MCC Lab', it does not specify any particular hardware details such as GPU models or CPU types.
Software Dependencies No The paper does not provide specific version numbers for software dependencies such as Python, PyTorch, or CUDA.
Experiment Setup Yes More experimental details can be found in the Appendix. ... Did you specify all the training details (e.g., data splits, hyperparameters, how they were chosen)? [Yes] See Aappendix C