How re-training process affect the performance of no-reference image quality metric for face images.
Xinwei LiuChristophe CharrierMarius PedersenPatrick BoursPublished in: Media Watermarking, Security, and Forensics (2019)
Keyphrases
- training process
- face images
- image quality assessment
- image quality metrics
- quality metrics
- image quality
- quality assessment
- face recognition
- training data
- structural similarity
- human faces
- facial expressions
- principal component analysis
- human visual system
- feature vectors
- low resolution
- subjective evaluation
- training set
- facial features
- image database
- perceptual quality
- support vector machine
- gabor wavelets
- test images
- high resolution
- input image
- neural network
- face verification
- visual quality
- learning process
- kullback leibler divergence
- natural images
- high quality
- correlation coefficient
- quality evaluation
- feature extraction
- adaboost algorithm
- image data
- visual perception
- human perception
- support vector
- quality attributes