-
I couldn't download the voxceleb1 identification split from this link https://mm.kaist.ac.kr/datasets/voxceleb. It only provides verification split. Does anyone know how to get the identification spli…
-
I'm encountering an issue when trying to use the CUDA provider with sherpa-rs on Windows.
I use the latest versions of CUDA, ONNX, and other dependencies. On my machine, CUDA works fine with my o…
-
调整了一下speakerlab/bin/infer_sv.py的推理代码,使用的模型是[https://www.modelscope.cn/models/iic/speech_eres2net_large_sv_en_voxceleb_16k/files](url),测试时听起来非常像的两个音频,其说话人相似度偏低。以0_s.wav和0_t.wav为例,取前3s的音频测试的相似度最高,为0.177…
-
我使用3dspeaker_speech_eres2net_base_sv_zh-cn_3dspeaker_16k.onnx做说话人识别
但是使用方式必须要提前拿几段数据先做声纹注册之后才能识别的比较准确。
但是在很多实际使用场景,是不可能提前做声纹注册的。
都是直接用一条长语音,含有多人说话的录音文件进行识别的。
测试3dspeaker_speech_eres2net_base_sv_zh…
-
Thank you for your well design of the ERes2Net model and make it open-source.
As you mention, the V2 version of ERes2Net improves the short-duration feature extraction capability of ERes2Net.
Are …
-
如下图所示,使用speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch进行语音识别,配合使用speech_eres2net_sv_zh-cn_16k-common作为说话人识别,加载speech_eres2net_sv_zh-cn_16k-common时报错了,是还不支持吗?能否支持啊?
![image](…
-
I tried two URL for downloding but give Connection timeout
**First url**
`https://mm.kaist.ac.kr/datasets/voxceleb`
**Second url**
`http://cnode01.mm.kaist.ac.kr/voxceleb/`
Where can I fi…
-
配置:win10
bat文件:
@echo off
CHCP 65001
set PYTHONIOENCODING=utf-8
python test.py ^
--silero-vad-model=./sherpa-onnx/silero_vad.onnx ^
--threshold=0.2 ^
--model=./sherpa-onnx/3dspeaker_spee…
-
I execute the command "python ./dataprep.py --save_path data --download --user USERNAME --password PASSWORD" with my own username and password but got the response below. And I also try to open the …
-
The link requesting USERNAME and PASSWORD doesn't work.
https://mm.kaist.ac.kr/datasets/voxceleb
How can I get the USERNAME and PASSWORD now?
Anyone can help me ?
I would really appreciate it.