Towards speech-to-text translation without speech recognition

TL;DR

This paper introduces a novel speech-to-text translation method that bypasses traditional speech recognition by using unsupervised pattern discovery and machine translation, enabling translation in low-resource, multi-speaker scenarios.

Contribution

It presents the first system for direct speech-to-text translation without ASR, utilizing unsupervised term discovery and a simple MT model on a realistic multi-speaker dataset.

Findings

System can correctly translate some content words

Low recall due to cross-speaker UTD challenges

Demonstrates feasibility of direct speech translation without recognition

Abstract

We explore the problem of translating speech to text in low-resource scenarios where neither automatic speech recognition (ASR) nor machine translation (MT) are available, but we have training data in the form of audio paired with text translations. We present the first system for this problem applied to a realistic multi-speaker dataset, the CALLHOME Spanish-English speech translation corpus. Our approach uses unsupervised term discovery (UTD) to cluster repeated patterns in the audio, creating a pseudotext, which we pair with translations to create a parallel text and train a simple bag-of-words MT model. We identify the challenges faced by the system, finding that the difficulty of cross-speaker UTD results in low recall, but that our system is still able to correctly translate some content words in test data.