OpenSalicon: An Open Source Implementation of the Salicon Saliency Model

TL;DR

This paper introduces OpenSalicon, an open-source implementation of the SALICON saliency model, facilitating research by providing training, testing, and pre-trained models for visual attention prediction.

Contribution

It provides the first publicly available implementation of SALICON, supporting training, testing, and pre-trained models to advance saliency research.

Findings

SALICON is among the top models on MIT 300 dataset

OpenSalicon enables easy benchmarking and extension

Pre-trained models facilitate quick saliency map generation

Abstract

In this technical report, we present our publicly downloadable implementation of the SALICON saliency model. At the time of this writing, SALICON is one of the top performing saliency models on the MIT 300 fixation prediction dataset which evaluates how well an algorithm is able to predict where humans would look in a given image. Recently, numerous models have achieved state-of-the-art performance on this benchmark, but none of the top 5 performing models (including SALICON) are available for download. To address this issue, we have created a publicly downloadable implementation of the SALICON model. It is our hope that our model will engender further research in visual attention modeling by providing a baseline for comparison of other algorithms and a platform for extending this implementation. The model we provide supports both training and testing, enabling researchers to quickly fine-tune the model on their own dataset. We also provide a pre-trained model and code for those users who only need to generate saliency maps for images without training their own model.