Visual aesthetic analysis using deep neural network: model and techniques to increase accuracy without transfer learning
Computer Vision and Pattern Recognition
2018-02-01 v4
Abstract
We train a deep Convolutional Neural Network (CNN) from scratch for visual aesthetic analysis in images and discuss techniques we adopt to improve the accuracy. We avoid the prevalent best transfer learning approaches of using pretrained weights to perform the task and train a model from scratch to get accuracy of 78.7% on AVA2 Dataset close to the best models available (85.6%). We further show that accuracy increases to 81.48% on increasing the training set by incremental 10 percentile of entire AVA dataset showing our algorithm gets better with more data.
Cite
@article{arxiv.1712.03382,
title = {Visual aesthetic analysis using deep neural network: model and techniques to increase accuracy without transfer learning},
author = {Muktabh Mayank Srivastava and Sonaal Kant},
journal= {arXiv preprint arXiv:1712.03382},
year = {2018}
}
Comments
Accepted at IEEE's 3rd International Conference for Convergence in Technology (I2CT) Pune - 7-8 April 2018