GuideSpeech & audio
Speech recognition: from wav2vec to Whisper
Two opposite routes to robust speech recognition: self-supervised pretraining on unlabeled audio (wav2vec 2.0) and weak supervision at web scale (Whisper).
Topic
Speech & audio
Guides, research, and posts on speech & audio, newest first.
All content on speech & audio
3 articles filed under Speech & audio
Content
PostImage & video
The Top 10 SI Innovations in Gaming and Virtual Worlds
Ten feature types in games and virtual worlds, from NPC dialogue to anti-cheat, with the player-trust questions a hype list skips.
PostImage & video
10 AI Art & Media Controversies, 2023 to 2025
Ten kinds of dispute around AI images, voice, music, and news video, framed as questions of rights and evidence rather than a gossip ranking.
Related: Speech & audio category: criteria and evaluation queue · All blog content