raghu@dark-factory :~/kb/streaming-asr $ cat

Streaming ASR

Automatic speech recognition that transcribes audio incrementally in real time, emitting partial and revisable (volatile) hypotheses before the utterance ends rather than decoding a completed clip in one batch pass.

grounded in: HN trend: on-device speech-to-text (Apple SpeechAnalyzer) benchmarked against Whisper — SpeechAnalyzer's defining property vs batch Whisper is streaming/real-time incremental transcription.

Connected concepts

On-device speech recognition, Time to First Token, In-browser inference

Explore it live in the knowledge graph →