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.emotion.whisper_emotion_dim import WhisperWrapper
5# Find device
6device = torch.device("cuda") if torch.cuda.is_available() else "cpu"
7# Load model from Huggingface
8model = WhisperWrapper.from_pretrained("tiantiaf/whisper-large-v3-msp-podcast-emotion-dim").to(device)
9model.eval()1# Load data, here just zeros as the example
2# Our training data filters output audio shorter than 3 seconds (unreliable predictions) and longer than 15 seconds (computation limitation)
3# So you need to prepare your audio to a maximum of 15 seconds, 16kHz and mono channel
4max_audio_length = 15 * 16000
5data = torch.zeros([1, 16000]).float().to(device)[:, :max_audio_length]
6arousal, valence, dominance = model(data)@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}
}