Enregistré dans:
Détails bibliographiques
Auteurs principaux: Mohammed, Fatma Youssef, Alexis, Kostas
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2506.02764
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866912411366195200
author Mohammed, Fatma Youssef
Alexis, Kostas
author_facet Mohammed, Fatma Youssef
Alexis, Kostas
contents Computational human attention modeling in free-viewing and task-specific settings is often studied separately, with limited exploration of whether a common representation exists between them. This work investigates this question and proposes a neural network architecture that builds upon the Human Attention transformer (HAT) to test the hypothesis. Our results demonstrate that free-viewing and visual search can efficiently share a common representation, allowing a model trained in free-viewing attention to transfer its knowledge to task-driven visual search with a performance drop of only 3.86% in the predicted fixation scanpaths, measured by the semantic sequence score (SemSS) metric which reflects the similarity between predicted and human scanpaths. This transfer reduces computational costs by 92.29% in terms of GFLOPs and 31.23% in terms of trainable parameters.
format Preprint
id arxiv_https___arxiv_org_abs_2506_02764
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Unified Attention Modeling for Efficient Free-Viewing and Visual Search via Shared Representations
Mohammed, Fatma Youssef
Alexis, Kostas
Computer Vision and Pattern Recognition
Artificial Intelligence
Computational human attention modeling in free-viewing and task-specific settings is often studied separately, with limited exploration of whether a common representation exists between them. This work investigates this question and proposes a neural network architecture that builds upon the Human Attention transformer (HAT) to test the hypothesis. Our results demonstrate that free-viewing and visual search can efficiently share a common representation, allowing a model trained in free-viewing attention to transfer its knowledge to task-driven visual search with a performance drop of only 3.86% in the predicted fixation scanpaths, measured by the semantic sequence score (SemSS) metric which reflects the similarity between predicted and human scanpaths. This transfer reduces computational costs by 92.29% in terms of GFLOPs and 31.23% in terms of trainable parameters.
title Unified Attention Modeling for Efficient Free-Viewing and Visual Search via Shared Representations
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2506.02764