This paper introduces Random Erasing, a novel data augmentation method for training convolutional neural networks (CNNs). Random Erasing randomly selects a rectangular region in an image and replaces its pixels with random values, generating images with varying levels of occlusion. This technique enhances the model's robustness to occlusion and reduces overfitting. The method is parameter-free, easy to implement, and can be integrated with most CNN-based models. It complements existing data augmentation techniques like random cropping and flipping, improving performance in image classification, object detection, and person re-identification tasks. The authors evaluate Random Erasing on several datasets, including CIFAR-10, CIFAR-100, Fashion-MNIST, PASCAL VOC 2007, Market-1501, DukeMTMC-reID, and CUHK03, demonstrating its effectiveness and robustness. The method consistently outperforms baselines and shows promise for various recognition tasks.This paper introduces Random Erasing, a novel data augmentation method for training convolutional neural networks (CNNs). Random Erasing randomly selects a rectangular region in an image and replaces its pixels with random values, generating images with varying levels of occlusion. This technique enhances the model's robustness to occlusion and reduces overfitting. The method is parameter-free, easy to implement, and can be integrated with most CNN-based models. It complements existing data augmentation techniques like random cropping and flipping, improving performance in image classification, object detection, and person re-identification tasks. The authors evaluate Random Erasing on several datasets, including CIFAR-10, CIFAR-100, Fashion-MNIST, PASCAL VOC 2007, Market-1501, DukeMTMC-reID, and CUHK03, demonstrating its effectiveness and robustness. The method consistently outperforms baselines and shows promise for various recognition tasks.