arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1803.09588cs.CV
收藏

用于预测深度学习精度的高效图像数据集分类难度估计

Efficient Image Dataset Classification Difficulty Estimation for Predicting Deep-Learning Accuracy

ETH Zürich(苏黎世联邦理工学院);IBM Research - Zürich(IBM苏黎世研究院);Queen’s University of Belfast(贝尔法斯特女王大学);Università di Bologna(博洛尼亚大学)

Florian Scheidegger, Roxana Istrate, Giovanni Mariani, Luca Benini, Costas Bekas, Cristiano Malossi

摘要:

在深度学习社区,新算法以令人难以置信的速度发布。因此,解决新数据集的图像分类问题成为一项具有挑战性的任务,因为它需要重新评估已发布的算法及其不同的配置,以便找到接近最佳的分类器。为了促进这一过程,在将我们的决策偏向于一类神经网络或在网络空间上运行昂贵的搜索之前,我们建议估计数据集的分类难度。我们的方法计算一个数字,它表征数据集的难度,比训练最先进的网络快27倍。所提出的方法可以与网络拓扑结构和超参数搜索优化器相结合,以有效地驱动搜索到有前途的神经网络配置。

英文摘要:

In the deep-learning community new algorithms are published at an incredible pace. Therefore, solving an image classification problem for new datasets becomes a challenging task, as it requires to re-evaluate published algorithms and their different configurations in order to find a close to optimal classifier. To facilitate this process, before biasing our decision towards a class of neural networks or running an expensive search over the network space, we propose to estimate the classification difficulty of the dataset. Our method computes a single number that characterizes the dataset difficulty 27x faster than training state-of-the-art networks. The proposed method can be used in combination with network topology and hyper-parameter search optimizers to efficiently drive the search towards promising neural-network configurations.

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。

↑