Verified75% confidenceFactTime unknown
In traditional voice interaction systems, paralinguistic information carried by speech—such as emotion, tone, and speech rate—is completely lost when converted to plain text during the ASR stage.
5
Sources
75%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
Related Entities
Related Claims
Unverified语音被转换为文字中间表示时，原始音频中约40%的信息随之消失，包括情绪、强调、语速等副语言信息83% similarUnverifiedMainstream voice assistance devices on the market can convert text or lip movements into sound, but cannot convey the speaker's genuine emotional inflections.78% similarUnverifiedASR systems can produce hallucinations, in which the output is fluent and grammatically correct yet entirely unrelated to the actual speech content of the input.76% similarVerifiedASR is a technology that converts human speech into text, and its core architecture has evolved from Gaussian Mixture Model–Hidden Markov Model (GMM-HMM) to end-to-end deep learning models.74% similarUnverifiedStreaming speaker diarization systems must output results in real time without access to the complete audio context and cannot retroactively correct already-output speaker labels.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/17980API
curl https://kongchang.com/api/v1/knowledge/claims/17980MCP
get_claim(id=17980)