Patent Leaderboard
USPTO Patent Rankings Data through Dec 31, 2025
Home › Best path change rate for unsupervised language model weight selection

Best path change rate for unsupervised language model weight selection

US Patent 10923117 · Granted Feb 16, 2021

Assignee

Inventors

View full patent text on Google Patents →

More Patents by These Inventors

Kalmannet: a learnable Kalman filter for acoustic echo cancellation2025
Efficient zero shot event extraction with context definition2025
Techniques for disentangled variational speech representation learning for zero-2025
Salience-aware cross-attention for abstractive summarization2025
Neuralecho: a self-attentive recurrent neural network for unified acoustic echo 2025
Learning singing from speech2025
Learning singing from speech2025
Multi-band synchronized neural vocoder2025
Extractive method for speaker identification in texts with self-training2025
Neural-echo: an unified deep neural network model for acoustic echo cancellation2025
End-to-end streaming speech translation with neural transducer2025
Progressive contrastive learning framework for self-supervised speaker verificat2025