| Author: | Zhou, Junlong |
| Title: | Video quality assessment using deep learning : a spatiotemporal approach |
| Advisors: | Chan, Yui Lam (EEE) |
| Degree: | M.Sc. |
| Year: | 2023 |
| Department: | Department of Electrical and Electronic Engineering |
| Pages: | 1 volume (unpaged) : color illustrations |
| Language: | English |
| Abstract: | This dissertation presents a novel approach to video quality assessment (VQA) based on two deep learning techniques: Transformers and 3D-convolutional neural networks (3D-CNNs). The primary focus is on using spatiotemporal learning for effective VQA. The first part of the study looks at the challenges associated with VQA and emphasizes the need for models that can accurately capture both temporal and spatial information in video sequences. To address these problems, we propose a unique framework that blends 3D-CNNs with Transformers. The 3D-CNNs are used to extract spatiotemporal features from the multi-frame by grouping the films into three-frame segments to extract short-term features. The long-term features included inside these short-term features are subsequently extracted using the Transformer model. Our proposed model exceeds the state-of-the-art methods already in use and performs well in the VQA challenge. The results show the potential of deep learning approaches in this field as well as the importance of integrating spatiotemporal learning into VQA. |
| Rights: | All rights reserved |
| Access: | restricted access |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| 8273.pdf | For All Users (off-campus access for PolyU Staff & Students only) | 1.67 MB | Adobe PDF | View/Open |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/13870