标签:

• 模型中心混合
• PyTorch模型中心混合 许可证: Apache-2.0 语言:
• 英语 评估指标:
• 准确率 基础模型:
• openai/whisper-large-v3 管道标签: 音频分类

Whisper Large v3 语音流畅度分类模型

模型描述

该模型实现了《Vox-Profile: 用于表征多样化说话者和语音特征的语音基础模型基准》(https://arxiv.org/pdf/2505.14648)中描述的语音流畅度分类方法。

模型首先以3秒窗口大小和1秒步长预测语音属于：

["流畅", "不流畅"]

若检测到不流畅语音，则进一步预测不流畅类型：

[
  "阻塞", 
  "延长", 
  "声音重复", 
  "词语重复", 
  "插入语"
]

使用方法

下载仓库

git clone git@github.com:tiantiaf0627/vox-profile-release.git

安装包

conda create -n vox_profile python=3.8
cd vox-profile-release
pip install -e .

加载模型

# 导入库
import torch
import torch.nn.functional as F
from src.model.fluency.whisper_fluency import WhisperWrapper

# 确定设备
device = torch.device("cuda") if torch.cuda.is_available() else "cpu"

# 从Huggingface加载模型
model = WhisperWrapper.from_pretrained("tiantiaf/whisper-large-v3-speech-flow").to(device)
model.eval()

预测准备

audio_data = torch.zeros([1, 16000*10]).float().to(device)
audio_segment = (audio_data.shape[1] - 3*16000) // 16000 + 1
if audio_segment < 1: audio_segment = 1
input_audio = list()
input_audio_length = list()
for idx in range(audio_segment): 
    input_audio.append(audio_data[0, 16000*idx:16000*idx+3*16000])
    input_audio_length.append(torch.tensor(len(audio_data[0, 16000*idx:16000*idx+3*16000])))
input_audio = torch.stack(input_audio, dim=0)
input_audio_length = torch.stack(input_audio_length, dim=0)

执行预测

fluency_outputs, disfluency_type_outputs = model(input_audio, length=input_audio_length)
fluency_prob   = F.softmax(fluency_outputs, dim=1).detach().cpu().numpy().astype(float).tolist()

disfluency_type_prob = nn.Sigmoid()(disfluency_type_outputs)
# 实际应用中可设置更高阈值
disfluency_type_predictions = (disfluency_type_prob > 0.7).int().detach().cpu().numpy().tolist()
disfluency_type_prob = disfluency_type_prob.cpu().numpy().astype(float).tolist()

汇总预测结果

utterance_fluency_list = list()
utterance_disfluency_list = list()
for audio_idx in range(audio_segment):
  disfluency_type = list()
  if fluency_prob[audio_idx][0] > 0.5: 
      utterance_fluency_list.append("流畅")
  else: 
      # 若预测为不流畅，则判断具体类型
      utterance_fluency_list.append("不流畅")
      predictions = disfluency_type_predictions[audio_idx]
      for label_idx in range(len(predictions)):
          if predictions[label_idx] == 1:
            disfluency_type.append(disfluency_type_labels[label_idx])
  utterance_disfluency_list.append(disfluency_type)

# 输出语句流畅度分析
print(utterance_fluency_list)
print(utterance_disfluency_list)

如有疑问，请联系：Tiantian Feng (tiantiaf@usc.edu)

若使用本模型或认为其有价值，请引用我们的论文

@article{feng2025vox,
  title={Vox-Profile: 用于表征多样化说话者和语音特征的语音基础模型基准},
  author={Feng, Tiantian and Lee, Jihwan and Xu, Anfeng and Lee, Yoonjeong and Lertpetchpun, Thanathai and Shi, Xuan and Wang, Helin and Thebaud, Thomas and Moro-Velazquez, Laureano and Byrd, Dani and others},
  journal={arXiv preprint arXiv:2505.14648},
  year={2025}
}