Source Separation of Convolutive and Noisy Mixtures Using Audio-Visual Dictionary Learning and Probabilistic Time-Frequency Masking.
Qingju LiuWenwu WangPhilip J. B. JacksonMark BarnardJosef KittlerJonathon A. ChambersPublished in: IEEE Trans. Signal Process. (2013)
Keyphrases
- generative model
- source separation
- audio visual
- dictionary learning
- audio features
- sparse representation
- multi modal
- blind source separation
- probabilistic model
- independent component analysis
- visual information
- semi supervised
- signal processing
- visual data
- denoising
- feature set
- frequency domain
- low level
- data sets
- visual features
- image representation
- data analysis
- image processing
- image classification
- high dimensional