Evaluation of pre-training impact on fine-tuning for remote sensing scene classification
Authors:
- 25 October 2018
- journal article
- research article
- Published by Taylor & Francis in Remote Sensing Letters
- Vol. 10 (1) , 49-58
Abstract
Fine-tuning a Deep Convolutional Neural Network (DCNN) model obtained from complicated and tedious pre-training on large-scale datasets seems to have become a standard baseline for various tasks with scale-limited datasets. However, there is little work done to investigate into the pre-training’s impact on fine-tuning. In this letter, we systematically verify that pre-training’s impact originating from large-scale datasets in other domain is still insurmountable for full-training with mainstream datasets in the Remote Sensing Scene Classification (RSSC) domain. However, we confirm that this impact is task-related. That is, once the need of pre-training is satisfied, excessive pre-training may lead to generalization degradation of following fine-tuning. This counter-intuitive finding explains why identical pre-training results in different generalization disparity in various tasks. We also provide experimental evidence to show that excessive or noisy pre-training is possible to make fine-tuning perform worse than full-training in some RSSC tasks. It is worth noticing that possible factors, including dataset, initialization, optimization and domain gap are taken into consideration, and more enlightening conclusions are given through this evaluation.Keywords
This publication has 13 references indexed in Scilit:
- Remote Sensing Image Scene Classification: Benchmark and State of the ArtProceedings of the IEEE, 2017
- Scene classification of high resolution remote sensing images using convolutional neural networks2016
- Deep Residual Learning for Image Recognition2016
- Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification2015
- Deep Learning Based Feature Selection for Remote Sensing Scene ClassificationIEEE Geoscience and Remote Sensing Letters, 2015
- Do deep features generalize from everyday objects to remote sensing and aerial scenes domains?2015
- Going deeper with convolutions2015
- Caffe2014
- Bag-of-visual-words and spatial extensions for land-use classification2010
- Flat MinimaNeural Computation, 1997