NOMAD: Unsupervised Learning of Perceptual Embeddings For Speech Enhancement and Non-Matching Reference Audio Quality Assessment.
Alessandro RaganoJan SkoglundAndrew HinesPublished in: ICASSP (2024)
Keyphrases
- quality assessment
- unsupervised learning
- reduced reference
- human visual system
- speech enhancement
- image quality assessment
- image quality
- visual information
- dimensionality reduction
- noisy environments
- video quality
- signal to noise ratio
- noise reduction
- perceptual image quality
- single channel
- human perception
- signal processing
- data quality
- visual quality
- speech signal
- expectation maximization
- feature selection
- wavelet domain
- visual data
- language model
- multiresolution
- linear prediction
- smoothing algorithm
- feature extraction
- machine learning