The Visual Saliency Transformer Goes Temporal: TempVST for Video Saliency Prediction

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F24%3A00379847" target="_blank" >RIV/68407700:21230/24:00379847 - isvavai.cz</a>

Výsledek na webu

<a href="https://doi.org/10.1109/ACCESS.2024.3436585" target="_blank" >https://doi.org/10.1109/ACCESS.2024.3436585</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/ACCESS.2024.3436585" target="_blank" >10.1109/ACCESS.2024.3436585</a>

Jazyk výsledku

angličtina

Název v původním jazyce

The Visual Saliency Transformer Goes Temporal: TempVST for Video Saliency Prediction

Popis výsledku v původním jazyce

The Transformer revolutionized Natural Language Processing and Computer Vision by effectively capturing contextual relationships in sequential data through its attention mechanism. While Transformers have been explored sufficiently in traditional computer vision tasks such as image classification, their application to more intricate tasks, such as Video Saliency Prediction (VSP), remains limited. Video saliency prediction is the task of identifying the most visually salient regions in a video, which are likely to capture a viewer's attention. In this study, we propose a pure transformer architecture named Temporal Visual Saliency Transformer (TempVST) for the VSP task. Our model leverages the Visual Saliency Transformer (VST) as a backbone, with the addition of a Transformer-based temporal module that can seamlessly transition diverse architectural frameworks from image to video domain, through the incorporation of temporal recurrences. Moreover, we demonstrate that transfer learning is viable in the context of VSP through Transformer architectures and helps reduce the duration of the training phase, leading to a reduction in the duration of the training phase by 41% and 45% in two different datasets. Our experiments were conducted on two benchmark datasets, DHF1K and LEDOV, and our results show that our network can compete with all other state-of-the-art models.

Název v anglickém jazyce

The Visual Saliency Transformer Goes Temporal: TempVST for Video Saliency Prediction

Popis výsledku anglicky

The Transformer revolutionized Natural Language Processing and Computer Vision by effectively capturing contextual relationships in sequential data through its attention mechanism. While Transformers have been explored sufficiently in traditional computer vision tasks such as image classification, their application to more intricate tasks, such as Video Saliency Prediction (VSP), remains limited. Video saliency prediction is the task of identifying the most visually salient regions in a video, which are likely to capture a viewer's attention. In this study, we propose a pure transformer architecture named Temporal Visual Saliency Transformer (TempVST) for the VSP task. Our model leverages the Visual Saliency Transformer (VST) as a backbone, with the addition of a Transformer-based temporal module that can seamlessly transition diverse architectural frameworks from image to video domain, through the incorporation of temporal recurrences. Moreover, we demonstrate that transfer learning is viable in the context of VSP through Transformer architectures and helps reduce the duration of the training phase, leading to a reduction in the duration of the training phase by 41% and 45% in two different datasets. Our experiments were conducted on two benchmark datasets, DHF1K and LEDOV, and our results show that our network can compete with all other state-of-the-art models.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2024

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

IEEE Access

ISSN

2169-3536

e-ISSN

2169-3536

Svazek periodika

12

Číslo periodika v rámci svazku

Aug

Stát vydavatele periodika

US - Spojené státy americké

Počet stran výsledku

12

Strana od-do

129705-129716

Kód UT WoS článku

001320453600001

EID výsledku v databázi Scopus

2-s2.0-85200252622