This thesis presents the development of a Graphical User Interface (GUI) in MATLAB for conducting a subjective image quality assessment using the pairwise comparison method. The objective of the study is to evaluate images generated by a Generative AI model based on human perception. The custom-designed GUI allows users to view and compare pairs of AI-generated images and express preferences through an intuitive interface. The collected responses are then analyzed to determine the perceived visual quality and ranking of the images. This approach provides insights into the subjective effectiveness of generative models and supports the development of more perceptually aligned AI image synthesis techniques.
This thesis presents the development of a Graphical User Interface (GUI) in MATLAB for conducting a subjective image quality assessment using the pairwise comparison method. The objective of the study is to evaluate images generated by a Generative AI model based on human perception. The custom-designed GUI allows users to view and compare pairs of AI-generated images and express preferences through an intuitive interface. The collected responses are then analyzed to determine the perceived visual quality and ranking of the images. This approach provides insights into the subjective effectiveness of generative models and supports the development of more perceptually aligned AI image synthesis techniques.
‘Quality assessment of Generative AI-based compression’
AMIRI MANESH, SANA
2024/2025
Abstract
This thesis presents the development of a Graphical User Interface (GUI) in MATLAB for conducting a subjective image quality assessment using the pairwise comparison method. The objective of the study is to evaluate images generated by a Generative AI model based on human perception. The custom-designed GUI allows users to view and compare pairs of AI-generated images and express preferences through an intuitive interface. The collected responses are then analyzed to determine the perceived visual quality and ranking of the images. This approach provides insights into the subjective effectiveness of generative models and supports the development of more perceptually aligned AI image synthesis techniques.| File | Dimensione | Formato | |
|---|---|---|---|
|
Amirimanesh_Sana.pdf
accesso aperto
Dimensione
16.89 MB
Formato
Adobe PDF
|
16.89 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/90289