Unispeech Sat Base 100h Libri Ft

🚀 UniSpeech-SAT-Base-Finetuned-100h-Libri

This is a unispeech-sat-base model that has been fine-tuned on 100 hours of Librispeech with 16kHz sampled speech audio. When using this model, ensure that your speech input is also sampled at 16kHz.

✨ Features

• Based on Microsoft's UniSpeech.
• Fine-tuned on 100 hours of LibriSpeech.
• Paper: UNISPEECH-SAT: UNIVERSAL SPEECH REPRESENTATION LEARNING WITH SPEAKER AWARE PRE-TRAINING

📦 Installation

No specific installation steps are provided in the original README.

💻 Usage Examples

Basic Usage

from transformers import Wav2Vec2Processor, UniSpeechSatForCTC
from datasets import load_dataset
import torch

# load model and tokenizer
processor = Wav2Vec2Processor.from_pretrained("microsoft/unispeech-sat-base-100h-libri-ft")
model = UniSpeechSatForCTC.from_pretrained("microsoft/unispeech-sat-base-100h-libri-ft")

# load dummy dataset
ds = load_dataset("patrickvonplaten/librispeech_asr_dummy", "clean", split="validation")

# tokenize
input_values = processor(ds[0]["audio"]["array"], return_tensors="pt", padding="longest").input_values  # Batch size 1

# retrieve logits
logits = model(input_values).logits

# take argmax and decode
predicted_ids = torch.argmax(logits, dim=-1)
transcription = processor.batch_decode(predicted_ids)

📚 Documentation

Abstract Self-supervised learning (SSL) is a long-standing goal for speech processing, since it utilizes large-scale unlabeled data and avoids extensive human labeling. Recent years witness great successes in applying self-supervised learning in speech recognition, while limited exploration was attempted in applying SSL for modeling speaker characteristics. In this paper, we aim to improve the existing SSL framework for speaker representation learning. Two methods are introduced for enhancing the unsupervised speaker information extraction. First, we apply the multi-task learning to the current SSL framework, where we integrate the utterance-wise contrastive loss with the SSL objective function. Second, for better speaker discrimination, we propose an utterance mixing strategy for data augmentation, where additional overlapped utterances are created unsupervisely and incorporate during training. We integrate the proposed methods into the HuBERT framework. Experiment results on SUPERB benchmark show that the proposed system achieves state-of-the-art performance in universal representation learning, especially for speaker identification oriented tasks. An ablation study is performed verifying the efficacy of each proposed method. Finally, we scale up training dataset to 94 thousand hours public audio data and achieve further performance improvement in all SUPERB tasks.

Authors: Sanyuan Chen, Yu Wu, Chengyi Wang, Zhengyang Chen, Zhuo Chen, Shujie Liu, Jian Wu, Yao Qian, Furu Wei, Jinyu Li, Xiangzhan Yu

🔧 Technical Details

The model was fine-tuned on 100 hours of LibriSpeech at 16kHz sampled speech audio.

📄 License

The official license can be found here

Additional Information

Widget Examples

• Librispeech sample 1
• Librispeech sample 2

Contribution

The model was contributed by cywang and patrickvonplaten.