You are here

Simple vs complex temporal recurrences for video saliency prediction

Authors: 

Panagiotis Linardos, Eva Mohedano, Juan Jose Nieto, Noel O'Connor, Xavier Giro-i-Nieto, Kevin McGuinness

Publication Type: 
Refereed Original Article
Abstract: 
This paper investigates modifying an existing neural network architecture for static saliency prediction using two types of recurrences that integrate information from the temporal domain. The first modification is the addition of a ConvLSTM within the architecture, while the second is a computationally simple exponential moving average of an internal convolutional state. We use weights pre-trained on the SALICON dataset and fine-tune our model on DHF1K. Our results show that both modifications achieve state-of-the-art results and produce similar saliency maps.
Digital Object Identifer (DOI): 
10.NA
Publication Status: 
Published
Date Accepted for Publication: 
Wednesday, 3 July, 2019
Publication Date: 
03/07/2019
Journal: 
arXiv preprint
Research Group: 
Institution: 
Dublin City University (DCU)
Open access repository: 
Yes