Picture for Bhuvana Ramabhadran

Bhuvana Ramabhadran

Speculative Speech Recognition by Audio-Prefixed Low-Rank Adaptation of Language Models

Add code
Jul 05, 2024
Viaarxiv icon

Speech Prefix-Tuning with RNNT Loss for Improving LLM Predictions

Add code
Jun 20, 2024
Viaarxiv icon

ASTRA: Aligning Speech and Text Representations for Asr without Sampling

Add code
Jun 10, 2024
Viaarxiv icon

Text Injection for Neural Contextual Biasing

Add code
Jun 05, 2024
Viaarxiv icon

Extending Multilingual Speech Synthesis to 100+ Languages without Transcribed Data

Add code
Feb 29, 2024
Viaarxiv icon

O-1: Self-training with Oracle and 1-best Hypothesis

Add code
Aug 14, 2023
Figure 1 for O-1: Self-training with Oracle and 1-best Hypothesis
Figure 2 for O-1: Self-training with Oracle and 1-best Hypothesis
Figure 3 for O-1: Self-training with Oracle and 1-best Hypothesis
Figure 4 for O-1: Self-training with Oracle and 1-best Hypothesis
Viaarxiv icon

Using Text Injection to Improve Recognition of Personal Identifiers in Speech

Add code
Aug 14, 2023
Figure 1 for Using Text Injection to Improve Recognition of Personal Identifiers in Speech
Figure 2 for Using Text Injection to Improve Recognition of Personal Identifiers in Speech
Figure 3 for Using Text Injection to Improve Recognition of Personal Identifiers in Speech
Figure 4 for Using Text Injection to Improve Recognition of Personal Identifiers in Speech
Viaarxiv icon

Large-scale Language Model Rescoring on Long-form Data

Add code
Jun 13, 2023
Figure 1 for Large-scale Language Model Rescoring on Long-form Data
Figure 2 for Large-scale Language Model Rescoring on Long-form Data
Figure 3 for Large-scale Language Model Rescoring on Long-form Data
Figure 4 for Large-scale Language Model Rescoring on Long-form Data
Viaarxiv icon

Understanding Shared Speech-Text Representations

Add code
Apr 27, 2023
Figure 1 for Understanding Shared Speech-Text Representations
Figure 2 for Understanding Shared Speech-Text Representations
Figure 3 for Understanding Shared Speech-Text Representations
Figure 4 for Understanding Shared Speech-Text Representations
Viaarxiv icon

Robust Knowledge Distillation from RNN-T Models With Noisy Training Labels Using Full-Sum Loss

Add code
Mar 10, 2023
Figure 1 for Robust Knowledge Distillation from RNN-T Models With Noisy Training Labels Using Full-Sum Loss
Figure 2 for Robust Knowledge Distillation from RNN-T Models With Noisy Training Labels Using Full-Sum Loss
Figure 3 for Robust Knowledge Distillation from RNN-T Models With Noisy Training Labels Using Full-Sum Loss
Figure 4 for Robust Knowledge Distillation from RNN-T Models With Noisy Training Labels Using Full-Sum Loss
Viaarxiv icon