Ultra2Speech - A Deep Learning Framework for Formant Frequency Estimation and Tracking from Ultrasound Tongue Images.
Pramit SahaYadong LiuBryan GickSidney S. FelsPublished in: MICCAI (3) (2020)
Keyphrases
- deep learning
- speech recognition
- input image
- image classification
- multiple images
- ultrasound images
- test images
- object recognition
- probabilistic model
- generative model
- bounding box
- lighting conditions
- data mining
- speech signal
- object class
- mean shift
- mental models
- vocal tract
- partial occlusion
- visual tracking
- segmentation method
- unsupervised learning
- image features
- image retrieval
- machine learning