Learn about acoustic unit discovery techniques in automatic speech recognition through this 19-minute conference presentation from the Center for Language & Speech Processing at Johns Hopkins University. Explore methods for automatically identifying and extracting meaningful acoustic units from speech signals without prior linguistic knowledge, covering computational approaches that can discover phoneme-like units directly from audio data. Examine the theoretical foundations and practical applications of unsupervised learning techniques in speech processing, including clustering algorithms and statistical models used to segment continuous speech into discrete acoustic units. Understand how these discovery methods contribute to building speech recognition systems for under-resourced languages and improve our understanding of speech perception and production mechanisms.