Efficient Image Dataset Classification Difficulty Estimation for Predicting Deep-Learning Accuracy
Abstract
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.
Cite
@article{arxiv.1803.09588,
title = {Efficient Image Dataset Classification Difficulty Estimation for Predicting Deep-Learning Accuracy},
author = {Florian Scheidegger and Roxana Istrate and Giovanni Mariani and Luca Benini and Costas Bekas and Cristiano Malossi},
journal= {arXiv preprint arXiv:1803.09588},
year = {2018}
}