Conceptos Básicos
NR-IQAモデルの敵対的な強度を軽減するための正則化戦略が効果的であることを示す。
Estadísticas
"predicted score: 37.3"
"MOS: 30.9"
"predicted score: 56.9"
"RMSE and PLCC are indicators of prediction accuracy"
Citas
"The NT strategy results in a slight decrease in the performance of NR-IQA models on clean images."
"Models trained with the proposed NT strategy exhibit significant improvement in adversarial robustness against both white-box and black-box attacks."