MMSpeech: Multi-modal Multi-task Encoder-Decoder Pre-training for Speech Recognition

Open in new window