Unverified60% confidenceFactExact time
The new generation voice AI model takes audio directly as modal input, skipping the transcription step to achieve more natural emotional perception and intonation variation.
2
Sources
60%
Confidence
Long-term
Relevance
7/18/2026
First Seen
Sources
Related Claims
VerifiedThe new end-to-end voice model is trained directly at the audio level, without first converting speech to text, and can directly capture acoustic features and emotions.83% similarUnverified经过大规模音频-频谱图配对数据训练的AI模型不仅能从频谱图还原通用语音，还能捕捉说话人的音色特征、口音和情绪状态80% similarVerifiedThe new generation of voice interaction has shifted to an end-to-end multimodal model architecture that understands and generates directly at the speech signal level, compressing latency and preserving paralinguistic information.80% similarVerifiedThe technology works in conjunction with human dubbing completed by professional voice actors, rather than AI-generated voices; AI is only responsible for visual lip-sync adjustment.79% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/551987API
curl https://kongchang.com/api/v1/knowledge/claims/551987MCP
get_claim(id=551987)