modelscope/3D-Speaker

Issues

• 在使用多模态的speaker-diarization报错

  #77 opened 10 months ago
  1

• language identification training problem

  #76 opened 10 months ago
  2

• about cam

  #75 opened 10 months ago
  2

• 在运行speaker diarization中的run_audio.sh时发生错误

  #74 opened 10 months ago
  10

• Inference acceleration

  #73 opened 10 months ago
  5

• ValueError: need at least one array to stack

  #72 opened 10 months ago
  0

• ValueError: need at least one array to stack

  #71 opened 10 months ago
  1

• 准备cnceleb wav

  #68 opened 10 months ago
  4

• windows

  #67 opened 10 months ago
  1

• how to extract embeddings by ecapa or resnet

  #66 opened 10 months ago
  3

• fine-tune

  #65 opened 10 months ago
  3

• Problem with training part.

  #59 opened 10 months ago
  3

• Error occurred during "bash run.sh" for speaker diarization

  #58 opened a year ago
  4

• speaker-diarization需要哪个版本的Funasr，第六步无法输出

  #57 opened a year ago
  2

• How to compute the ERes2Net model param?

  #55 opened a year ago
  6

• Missing transcripts?

  #54 opened a year ago
  4

• Exmaple can not run

  #53 opened a year ago
  2

• 人声分离效果较差

  #52 opened a year ago
  14

• Could you provide the md5 value of train.tar.gz-part-{a-f} ?

  #49 opened a year ago
  7

• 请问可以分离出指定人的语音吗

  #48 opened a year ago
  1

• DDP WARNING

  #47 opened a year ago
  3

• damo/speech_eres2net_sv_zh-cn_16k-common 模型不支持输入numpy数组

  #46 opened a year ago
  1

• 错误数据

  #45 opened a year ago
  0

• 关于ERes2Net的250k新模型

  #44 opened a year ago
  1

• transcription

  #43 opened a year ago
  1

• The problem about the selection of num_of_spk in speaker-diarization

  #42 opened a year ago
  2

• 如何用这些个模型跑自己的数据集？

  #41 opened a year ago
  1

• [BUG] 提fbank的频率和重采样之后的频率不一致

  #40 opened a year ago
  2

• Can you consider release the ERes2Net-Base-200k-Spkrs model?

  #39 opened a year ago
  1

• EResNet result on VoxCeleb is not comparable

  #38 opened a year ago
  2

• 案例写的是比较两个人音频是否为同一个人，那如果与很多人的音频库比较怎么操作呢

  #37 opened a year ago
  11

• 运行bash run.sh 出错

  #36 opened a year ago
  4

• 关于 sv 识别结果的问题

  #35 opened a year ago
  2

• 请问为在训练CAM++模型时使用多卡训练显示Reducer buckets have been rebuilt in this iteration是正常的吗

  #34 opened a year ago
  1

• speech_eres2net_sv_zh-cn_16k-common预训练模型相关问题

  #32 opened a year ago
  1

• 请问cam++ 适合做文本相关的说话人确认任务吗？

  #30 opened a year ago
  3

• about the result

  #28 opened a year ago
  7

• Methods for fine-tuning of pretrained models in modelscope

  #27 opened a year ago
  1

• 关于learning rate

  #26 opened a year ago
  2

• cnceleb 训练配置问题

  #24 opened a year ago
  3

• 关于voxceleb dino

  #23 opened a year ago
  14

• 麻烦问下，3d speaker 开源数据里的标签代表什么意思？

  #22 opened a year ago
  2

• Apache-2.0 协议涵盖范围

  #19 opened a year ago
  2

• 分类器？

  #18 opened a year ago
  1

• 请问如果我要用CNceleb数据集训练cam++说话人验证模型，那我应该用哪个脚本

  #17 opened a year ago
  1

• is speech_campplus_speaker-diarization_common onnx model available?

  #16 opened a year ago
  6

• 我能问一下，前两天在达摩院重新上传的cam++的预训练模型与没重新上传之前的预训练模型有什么不一样吗，因为我在这个模型上加了一些模块，发现效果还没之前那个版本的好

  #15 opened a year ago
  2

• 我用达摩院提供的cam++预训练模型在vox1-o数据集上提取embedding计算eer是0.8666，而论文里是0.7几

  #14 opened a year ago
  3

• Inconsistent Performance and Loss when Resuming Training

  #13 opened a year ago
  4

• compute_score_metrics.py 输入参数？

  #12 opened a year ago
  3