Patent Leaderboard
USPTO Patent Rankings Data through Dec 31, 2025
Home › N-best softmax smoothing for minimum bayes risk training of attention based sequence-to-sequence models

N-best softmax smoothing for minimum bayes risk training of attention based sequence-to-sequence models

US Patent 11803618 · Granted Oct 31, 2023

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
System and method for motion prediction in autonomous driving2025
Neuralecho: a self-attentive recurrent neural network for unified acoustic echo 2025
Learning singing from speech2025
Learning singing from speech2025
Learning singing from speech2025
Multi-band synchronized neural vocoder2025
Multi-band synchronized neural vocoder2025
Extractive method for speaker identification in texts with self-training2025