Views
No views yet
git clone git@github.com:tiantiaf0627/vox-profile-release.gitconda create -n vox_profile python=3.8
cd vox-profile-release
pip install -e .1# Load libraries
2import torch
3import torch.nn.functional as F
4from src.model.accent.wavlm_accent import WavLMWrapper
5
6# Find device
7device = torch.device("cuda") if torch.cuda.is_available() else "cpu"
8
9# Load model from Huggingface
10model = WavLMWrapper.from_pretrained("tiantiaf/wavlm-large-broader-accent").to(device)
11model.eval()1# Label List
2english_accent_list = [
3 'British Isles', 'North America', 'Other'
4]
5
6# Load data, here just zeros as the example, audio data should be 16kHz mono channel
7data = torch.zeros([1, 16000]).float().to(device)
8logits, embeddings = model(data, return_feature=True)
9
10# Probability and output
11accent_prob = F.softmax(logits, dim=1)
12print(english_accent_list[torch.argmax(accent_prob).detach().cpu().item()])@article{feng2025vox,
title={Vox-Profile: A Speech Foundation Model Benchmark for Characterizing Diverse Speaker and Speech Traits},
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}
}