skxooo
skxooo
Would it be possible for you to share some sample data with me? You can send it to my email: [email protected].
老师您好，我想请问一下如何正确地生成CALLHOME数据集的spk2utt和utt2spk文件？我目前根据Kaldi Recipe的make_callhome.sh文件获得了full_rttm和segments文件、生成了utt2spk和spk2utt文件以及callhome1和callhome2子集，但是我发现生成的utt2spk文件和spk2utt文件并没有区分full_rttm文件中不同的说话人。比如：对于iaaa.sph **full_rttm中部分内容为：** SPEAKER iaaa 0 0 1.08 B SPEAKER iaaa 0 0.85 0.39 A SPEAKER iaaa 0 1.12 2.14 B SPEAKER iaaa 0 1.45 1.82 A SPEAKER iaaa 0...
您好，请问西湖大学是否购买了您在文章中使用的LDC公司数据集？我注意到文章中提到的数据集版权均归LDC公司所有，且使用这些数据集需要付费或加入会员。期待您的解答，谢谢！
How can I obtain the correct `spk2utt` and `utt2spk` files? I used the `kaldi/egs/callhome_diarization/v1/local/make_callhome.sh` script to generate the `spk2utt` and `utt2spk` files for CallHome, but it did not differentiate between...
The HMDB51-DVS dataset is extracted as **.aedat** files rather than **.mat** files, but the script **hmdb_dvs.py** loads the dataset in **.mat** format. How should this issue be addressed?