arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2007.07805cs.CV
收藏

使用数据增强、焦点余弦损失和增强的图像分类数据高效深度学习方法

Data-Efficient Deep Learning Method for Image Classification Using Data Augmentation, Focal Cosine Loss, and Ensemble

Zuminternet

Byeongjo Kim, Chanran Kim, Jaehoon Lee, Jein Song, Gyoungsoo Park

摘要:

一般来说,充足的数据对于深度学习模型的更好性能和泛化至关重要。然而,有很多限制(成本、资源等)数据汇集的不充分导致了大多数领域数据的缺乏。此外,每个数据源和许可证的不同领域也导致难以收集足够的数据。这种情况使得我们不仅难以利用预先训练的模型,而且难以利用外部知识。因此,有效地利用小数据集以实现更好的性能非常重要。我们在三个方面应用了一些技术:数据,损失函数和预测,以实现用更少的数据从头开始训练。通过这些方法,我们利用ImageNet数据获得了高准确性,每个类仅包含50张图像。此外,我们的模型在数据有效的计算机视觉挑战的视觉归纳中排名第四。

英文摘要:

In general, sufficient data is essential for the better performance and generalization of deep-learning models. However, lots of limitations(cost, resources, etc.) of data collection leads to lack of enough data in most of the areas. In addition, various domains of each data sources and licenses also lead to difficulties in collection of sufficient data. This situation makes us hard to utilize not only the pre-trained model, but also the external knowledge. Therefore, it is important to leverage small dataset effectively for achieving the better performance. We applied some techniques in three aspects: data, loss function, and prediction to enable training from scratch with less data. With these methods, we obtain high accuracy by leveraging ImageNet data which consist of only 50 images per class. Furthermore, our model is ranked 4th in Visual Inductive Printers for Data-Effective Computer Vision Challenge.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑