LEIBOLD RALF

Multi-encoder end-to-end automatic speech recognition (ASR) for joint modeling of multiple input devices