Investigation of Whisper ASR Hallucinations Induced by Non-Speech Audio · HackerPapers discussionshttps://hackerpaper.kaipfister.de/p/2501.11378/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMTInvestigation of Whisper ASR Hallucinations Induced by Non-Speech Audio — mentioned by rhdunnhttps://news.ycombinator.com/item?id=49371085urn:hackerpapers:mention:2501.11378:49371085Thu, 20 Aug 2026 06:27:00 GMT<p>Speech recognition is a notoriously difficult problem to solve. It can work great as a first pass that someone can fix up, but not on its own. This is why having sentence/clause and word-level transcription markers along side a recording help. One of the main issues is around homophones in an accent (Adam/Atom in American English, Bath/Barf in London English, etc.). Not to mention pronunciation variations due to fast speech, speech impedements, or parts of words side-by-side that sound like a different word. Another big issue is around misaligned training data. For example, Whisper is known…</p><p><a href="https://hackerpaper.kaipfister.de/p/2501.11378/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.SD