# Phonetically-Oriented Word Error Alignment for Speech Recognition Error Analysis in Speech Translation

## TL;DR

This paper introduces POWER, a phoneme-based extension to WER, enhancing alignment accuracy and capturing homophonic errors in speech recognition, thereby improving error analysis for speech translation.

## Contribution

It presents a novel phonetic alignment method that refines word error analysis by incorporating phoneme-level information without requiring time boundary data.

## Key findings

- POWER provides better word alignments than traditional WER.
- It effectively captures homophonic errors in speech recognition.
- The method improves error analysis for downstream speech translation tasks.

## Abstract

We propose a variation to the commonly used Word Error Rate (WER) metric for speech recognition evaluation which incorporates the alignment of phonemes, in the absence of time boundary information. After computing the Levenshtein alignment on words in the reference and hypothesis transcripts, spans of adjacent errors are converted into phonemes with word and syllable boundaries and a phonetic Levenshtein alignment is performed. The aligned phonemes are recombined into aligned words that adjust the word alignment labels in each error region. We demonstrate that our Phonetically-Oriented Word Error Rate (POWER) yields similar scores to WER with the added advantages of better word alignments and the ability to capture one-to-many word alignments corresponding to homophonic errors in speech recognition hypotheses. These improved alignments allow us to better trace the impact of Levenshtein error types on downstream tasks such as speech translation.