Skip to content

feat: add pronunciation test framework for multilingual TTS (closes #120)#139

Open
udayreddy-s wants to merge 2 commits into
sugarlabs:mainfrom
udayreddy-s:feat/pronunciation-test-framework
Open

feat: add pronunciation test framework for multilingual TTS (closes #120)#139
udayreddy-s wants to merge 2 commits into
sugarlabs:mainfrom
udayreddy-s:feat/pronunciation-test-framework

Conversation

@udayreddy-s
Copy link
Copy Markdown

Closes #120

I opened issue #120 and this is the implementation.

Adds tests/test_pronunciation.py with ground truth phoneme
test cases for 5 languages: Hindi, Spanish, French, Arabic,
and Swahili.

Key design decisions:

  • Hindi cases verified by me as a native speaker — covers
    schwa deletion (Ohala 1983) which is the most critical
    rule for natural-sounding Hindi TTS
  • Works in two modes: documentation mode (no TTS needed)
    and live mode (pass any tts_func that returns IPA output)
  • Easy to extend — just add new language dict to
    ALL_LANGUAGE_TESTS
  • No new dependencies required

@mebinthattil @Ibiam — happy to add more languages or
adjust the test format based on your feedback.

Relates to GSoC 2026 Speak-AI Multilingual Support proposal.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

feat: Add pronunciation test framework for multilingual TTS validation

1 participant