Relative Positional Encoding for Speech Recognition and Direct Translation
This work addresses the challenge of improving speech recognition and translation for applications like voice assistants and multilingual communication, though it is incremental as it builds on existing Transformer and relative encoding methods.
The authors tackled the problem of Transformer models using text-optimized positional encoding for speech tasks, which is suboptimal for acoustic inputs, by adapting relative position encoding to Speech Transformers, resulting in state-of-the-art performance on Switchboard speech recognition and MuST-C speech translation benchmarks.
Transformer models are powerful sequence-to-sequence architectures that are capable of directly mapping speech inputs to transcriptions or translations. However, the mechanism for modeling positions in this model was tailored for text modeling, and thus is less ideal for acoustic inputs. In this work, we adapt the relative position encoding scheme to the Speech Transformer, where the key addition is relative distance between input states in the self-attention network. As a result, the network can better adapt to the variable distributions present in speech data. Our experiments show that our resulting model achieves the best recognition result on the Switchboard benchmark in the non-augmentation condition, and the best published result in the MuST-C speech translation benchmark. We also show that this model is able to better utilize synthetic data than the Transformer, and adapts better to variable sentence segmentation quality for speech translation.