Dynamic Test-Time Augmentation via Differentiable Functions
Distribution shifts, which often occur in the real world, degrade the accuracy of deep learning systems, and thus improving robustness to distribution shifts is essential for practical applications. To improve robustness, we study an image enhancement method that generates recognition-friendly images without retraining the recognition model. We propose a novel image enhancement method, DynTTA, which is based on differentiable data augmentation techniques and generates a blended image from many augmented images to improve the recognition accuracy under distribution shifts. In addition to standard data augmentations, DynTTA also incorporates deep neural network-based image transformation, further improving the robustness. Because DynTTA is composed of differentiable functions, it can be directly trained with the classification loss of the recognition model. In experiments with widely used image recognition datasets using various classification models, DynTTA improves the robustness with almost no reduction in classification accuracy for clean images, thus outperforming the existing methods. Furthermore, the results show that robustness is significantly improved by estimating the training-time augmentations for distribution-shifted datasets using DynTTA and retraining the recognition model with the estimated augmentations. DynTTA is a promising approach for applications that require both clean accuracy and robustness. Our code is available at \url{https://github.com/s-enmt/DynTTA}.
Code (1)
Tasks
ClassificationData AugmentationImage EnhancementMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Differentiable Augmentation for Data-Efficient GAN Training
The performance of generative adversarial networks (GANs) heavily deteriorates given a limited amount of training data. This is mainly because the discriminator is memorizing the exact training set. To combat it, we prop…
Image GenerationEnd-to-End Augmentation Hyperparameter Tuning for Self-Supervised Anomaly Detection
Self-supervised learning (SSL) has emerged as a promising paradigm that presents self-generated supervisory signals to real-world problems, bypassing the extensive manual labeling burden. SSL is especially attractive for…
Anomaly DetectionData AugmentationSelf-Supervised Anomaly DetectionSelf-Supervised Learning+1Neural Weighted A*: Learning Graph Costs and Heuristics with Differentiable Anytime A*
Recently, the trend of incorporating differentiable algorithms into deep learning architectures arose in machine learning research, as the fusion of neural layers and algorithmic layers has been beneficial for handling c…
Invariance Learning in Deep Neural Networks with Differentiable Laplace Approximations
Data augmentation is commonly applied to improve performance of deep learning by enforcing the knowledge that certain transformations on the input preserve the output. Currently, the data augmentation parameters are chos…
Data AugmentationGaussian ProcessesModel SelectionDeep differentiable reinforcement learning and optimal trading
In many reinforcement learning applications, the underlying environment reward and transition functions are explicitly known differentiable functions. This enables us to use recent research which applies machine learning…
Portfolio Optimizationreinforcement-learningReinforcement LearningReinforcement Learning (RL)