Weakly-Supervised Speech Pre-training: A Case Study on Target Speech Recognition.
Wangyou ZhangYanmin QianPublished in: CoRR (2023)
Keyphrases
- speech recognition
- weakly supervised
- isolated word
- speech signal
- acoustic models
- speech synthesis
- object detectors
- automatic speech recognition
- speech recognizer
- speech processing
- superpixels
- hidden markov models
- language model
- object class
- topic models
- relation extraction
- speaker identification
- speaker independent
- speech recognition technology
- semi supervised
- pattern recognition
- noisy environments
- speech recognition systems
- speaker dependent
- training set
- named entities
- multiscale
- co occurrence
- image segmentation
- image features
- multi modal
- natural images
- computer vision
- bounding box