Verified65% confidenceFactTime unknown
Whisper基于Transformer编码器-解码器架构，在68万小时的多语言标注音频数据上进行弱监督训练
3
Sources
65%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedWhisper is an automatic speech recognition model open-sourced by OpenAI in 2022, built on a Transformer architecture and trained on 680,000 hours of multilingual audio data.82% similarUnverified不同语音识别模型架构差异显著：Whisper使用纯Transformer处理Mel频谱图，Meta的wav2vec 2.0和HuBERT依赖CNN特征提取器加自监督预训练，NVIDIA的Conformer系列将卷积模块嵌入Transformer层77% similarVerifiedModels such as Whisper and Wav2Vec 2.0 learn general acoustic representations on large amounts of unlabeled audio through self-supervised pre-training, reducing reliance on manually annotated data.74% similarUnverifiedSpeech synthesis models based on the Transformer architecture, pre-trained on massive multi-speaker corpora, can extract voiceprint features such as timbre, intonation, and rhythm from very short audio clips.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/23275API
curl https://kongchang.com/api/v1/knowledge/claims/23275MCP
get_claim(id=23275)