Despite several approaches to recover the ground truth subjective quality score from noisy individual ratings in subjective experiments have been explored in the literature, there is still room for improvement, in particular in terms of robustness to noise. This paper proposes a new approach that combines the traditional maximum likelihood estimation framework with a newly proposed regularization term, based on information theory concepts, that is meant to underweight surprising ratings of the quality of a given stimulus, looked at as a noise manifestation, in the final analytical expression of the recovered subjective quality. Computational experiments show the higher robustness to noise of our proposal when compared to three state-of-the-art methods.

Regularized Maximum Likelihood Estimation of the Subjective Quality from Noisy Individual Ratings / FOTIO TIOTSOP, Lohic; Servetti, Antonio; Barkowsky, Marcus; Masala, Enrico. - STAMPA. - (2022). (Intervento presentato al convegno The 14th International Conference on Quality of Multimedia Experience (QoMEX) tenutosi a Lippstadt (Germany) nel 5-7 Sept 2022) [10.1109/QoMEX55416.2022.9900903].

Regularized Maximum Likelihood Estimation of the Subjective Quality from Noisy Individual Ratings

Lohic Fotio Tiotsop;Antonio Servetti;Enrico Masala
2022

Abstract

Despite several approaches to recover the ground truth subjective quality score from noisy individual ratings in subjective experiments have been explored in the literature, there is still room for improvement, in particular in terms of robustness to noise. This paper proposes a new approach that combines the traditional maximum likelihood estimation framework with a newly proposed regularization term, based on information theory concepts, that is meant to underweight surprising ratings of the quality of a given stimulus, looked at as a noise manifestation, in the final analytical expression of the recovered subjective quality. Computational experiments show the higher robustness to noise of our proposal when compared to three state-of-the-art methods.
2022
9781665487948
File in questo prodotto:
File Dimensione Formato  
QoMEX2022_FotioTiotsop_final_submitted.pdf

accesso aperto

Descrizione: versione finale accettata dalla conferenza
Tipologia: 2. Post-print / Author's Accepted Manuscript
Licenza: PUBBLICO - Tutti i diritti riservati
Dimensione 679.89 kB
Formato Adobe PDF
679.89 kB Adobe PDF Visualizza/Apri
Regularized_Maximum_Likelihood_Estimation_of_the_Subjective_Quality_from_Noisy_Individual_Ratings.pdf

non disponibili

Tipologia: 2a Post-print versione editoriale / Version of Record
Licenza: Non Pubblico - Accesso privato/ristretto
Dimensione 721.09 kB
Formato Adobe PDF
721.09 kB Adobe PDF   Visualizza/Apri   Richiedi una copia
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11583/2971779