Unverified50% confidenceFactExact time
After the rise of deep learning post-2012, end-to-end neural network models reduced speech recognition error rates in general scenarios to below 5%.
1
Sources
50%
Confidence
Long-term
Relevance
9/2/2026
First Seen
Sources
Related Entities
Related Claims
Unverified2012年深度学习浪潮兴起后，DNN-HMM混合架构逐步取代纯GMM方案，此后CTC损失函数普及和端到端模型崛起带来语音识别范式革命70% similarUnverifiedSPLADE and DeepImpact are representatives of learned sparse retrieval, producing sparse vectors whose per-term weights are learned by a neural network and enabling term expansion.70% similarUnverifiedAfter the deep learning wave of 2012, deep neural networks gradually replaced GMMs to form a DNN-HMM hybrid architecture.69% similarUnverifiedElevenLabs基于深度学习模型，能够生成接近真人水平的语音输出，捕捉语调、节奏、情感等细微特征68% similarUnverified深度神经网络降噪的代表性方案包括RNNoise和DeepFilterNet68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/845839API
curl https://kongchang.com/api/v1/knowledge/claims/845839MCP
get_claim(id=845839)