Multi-Quartznet: Multi-Resolution Convolution for Speech Recognition with Multi-Layer Feature Fusion.
Jian LuoJianzong WangNing ChengGuilin JiangJing XiaoPublished in: SLT (2021)
Keyphrases
- speech recognition
- multi layer
- feature fusion
- multiresolution
- hidden markov models
- feature extraction
- multiple features
- neural network
- neural nets
- language model
- pattern recognition
- image processing
- probabilistic model
- wavelet transform
- wavelet domain
- object recognition
- high level
- multiscale
- image fusion
- active learning
- artificial neural networks
- computer vision