mirror of
https://github.com/cmusphinx/pocketsphinx.git
synced 2026-06-16 12:24:32 +00:00
Kevin Lenzo
7dffccf9dc
Update cmudict-en-us.dict from latest CMUdict source with proper handling of stress marker removal and duplicate elimination. Source: https://github.com/cmusphinx/cmudict Changes: - Updated from older CMUdict snapshot to current version - Strip stress markers (0, 1, 2) from all phonemes - Remove duplicate pronunciations created by stress normalization - Renumber pronunciation variants sequentially after deduplication - Strip trailing comments from dictionary entries Statistics: - Previous: 134,782 entries - Current: 134,860 entries (net +78 entries) - Removed: 306 duplicate pronunciations after stress removal - Added: 384 new words and variants from updated CMUdict Notable changes: - Many words updated with improved phonetic transcriptions - New place names and proper nouns added - Obsolete pronunciation variants removed - Words ending in -ism updated with correct schwa pronunciation The new dictionary maintains full compatibility with PocketSphinx while incorporating the latest lexical improvements from the CMUdict project.