论文标题
混合自我监督的学习用于对比度不足的应用
Mix-up Self-Supervised Learning for Contrast-agnostic Applications
论文作者
论文摘要
对比的自我监督学习最近引起了重大的研究关注。它通过将相同图像的增强视图彼此接近，同时推开不同图像的嵌入，从而从未标记的数据中学习有效的视觉表示。尽管在Imagenet分类，可可对象检测等方面取得了巨大的成功，但其性能在对比度不足的应用程序（例如医学图像分类）上却降低了，其中所有图像在视觉上彼此相似。由于图像之间的距离很小，因此在优化嵌入空间方面造成了困难。为了解决这个问题，我们提出了第一个用于对比度应用程序的自我监督学习框架。我们基于跨域混音解决了跨图像的较低差异，并基于两个协同目标来构建借口任务：图像重建和透明度预测。两个基准数据集的实验结果验证了我们方法的有效性，在该方法中，与现有的自我监督学习方法相比，获得了2.5％〜7.4％的TOP-1精度。
Contrastive self-supervised learning has attracted significant research attention recently. It learns effective visual representations from unlabeled data by embedding augmented views of the same image close to each other while pushing away embeddings of different images. Despite its great success on ImageNet classification, COCO object detection, etc., its performance degrades on contrast-agnostic applications, e.g., medical image classification, where all images are visually similar to each other. This creates difficulties in optimizing the embedding space as the distance between images is rather small. To solve this issue, we present the first mix-up self-supervised learning framework for contrast-agnostic applications. We address the low variance across images based on cross-domain mix-up and build the pretext task based on two synergistic objectives: image reconstruction and transparency prediction. Experimental results on two benchmark datasets validate the effectiveness of our method, where an improvement of 2.5% ~ 7.4% in top-1 accuracy was obtained compared to existing self-supervised learning methods.
