1 paper
Timo Lohrenz, Patrick Schwarz, Zhengyang Li +1
Recently, attention-based encoder-decoder (AED) models have shown high performance for end-to-end automatic speech recognition (ASR) across several tasks. Addressing overconfidence…