Acoustic or Pattern? Speech Spoofing Countermeasure based on Image Pre-training Models.
Jingze LuZhuo LiYuxiang ZhangWenchao WangPengyuan ZhangPublished in: DDAM@MM (2022)
Keyphrases
- acoustic models
- input image
- speech recognition
- single image
- image data
- image analysis
- image features
- multiscale
- image content
- image classification
- random fields
- image retrieval
- bayesian framework
- low level
- image segmentation
- test images
- image representation
- segmentation method
- edge detection
- broadcast news
- neural network
- gray level
- segmentation algorithm
- training examples
- language model
- region of interest
- color images
- image collections
- training set
- speech recognizer
- speech recognition systems