Datasets:

wsntxxn
/

DailyTalkEdit

text stringlengths 14 147
LA_E_4123318.flac
LA_E_9964383.flac
LA_E_2411171.flac
LA_E_9339758.flac
LA_E_2775552.flac
LA_E_1774443.flac
LA_E_6772958.flac
LA_E_2476686.flac
LA_E_2966780.flac
LA_E_3023308.flac
LA_E_8188227.flac
LA_E_5174081.flac
LA_E_5254567.flac
LA_E_4077147.flac
LA_E_4212023.flac
LA_E_2169669.flac
LA_E_3798060.flac
LA_E_9775814.flac
LA_E_6130485.flac
LA_E_8281985.flac
LA_E_1170060.flac
LA_E_1863583.flac
LA_E_4544319.flac
LA_E_8298224.flac
LA_E_1036359.flac
LA_E_1063956.flac
LA_E_4857219.flac
LA_E_2187761.flac
LA_E_6377807.flac
LA_E_3708023.flac
LA_E_2833287.flac
LA_E_5147021.flac
LA_E_7766134.flac
LA_E_1136035.flac
LA_E_8480587.flac
LA_E_8524198.flac
LA_E_8941036.flac
LA_E_7002917.flac
LA_E_9209859.flac
LA_E_5810486.flac
LA_E_2362820.flac
LA_E_3497058.flac
LA_E_3249698.flac
LA_E_2379834.flac
LA_E_4900217.flac
LA_E_4773224.flac
LA_E_6446352.flac
LA_E_1092011.flac
LA_E_2571890.flac
LA_E_1654041.flac
LA_E_5194837.flac
LA_E_8022053.flac
LA_E_4206334.flac
LA_E_8636666.flac
LA_E_1537449.flac
LA_E_1849845.flac
LA_E_6800434.flac
LA_E_9174637.flac
LA_E_3935560.flac
LA_E_8653975.flac
LA_E_9554032.flac
LA_E_6042049.flac
LA_E_1868065.flac
LA_E_2961372.flac
LA_E_8929484.flac
LA_E_1765365.flac
LA_E_8175756.flac
LA_E_5232598.flac
LA_E_2359084.flac
LA_E_8203627.flac
LA_E_1070406.flac
LA_E_9862485.flac
LA_E_5352599.flac
LA_E_9060445.flac
LA_E_6421546.flac
LA_E_6947941.flac
LA_E_8388075.flac
LA_E_6994204.flac
LA_E_1164778.flac
LA_E_4826417.flac
LA_E_1596483.flac
LA_E_7564751.flac
LA_E_8460538.flac
LA_E_9348205.flac
LA_E_4396063.flac
LA_E_6342725.flac
LA_E_9960399.flac
LA_E_7593523.flac
LA_E_4501281.flac
LA_E_8702425.flac
LA_E_4270860.flac
LA_E_2497555.flac
LA_E_7894885.flac
LA_E_2326699.flac
LA_E_9762899.flac
LA_E_1056197.flac
LA_E_7554303.flac
LA_E_3864228.flac
LA_E_4326961.flac
LA_E_8867208.flac

End of preview. Expand in Data Studio

Dataset Structure

This repo contains DailyTalkEdit and the extended semantic influence annotation of PartialEdit.

daily_talk_edit/
- concat_utts/: paired audio files per sample, obtained by concatenating all utterances in dialogues
  - <id>_original.wav
  - <id>_modified.wav
- modified_utts/: modified-only audio segments
  - <id>_<utt_id>_modified.wav
- metadata/
  - train.jsonl, val.jsonl, test.jsonl
    - filename: audio file name
    - fake_region: modified time ranges
    - semantic_influence: text description of semantic impact
  - modify_config_pipeline.json: modification details
partial_edit/
- metadata/semantic_influence.jsonl
  - utt_id: sample ID
  - influence: text description of semantic impact
sampled_test_data/
- *.txt: data IDs of the sampled data in the mixed evaluation set in the paper "HoliAntiSpoof: Audio LLM for Holistic Speech Anti-Spoofing".

Downloads last month: 6,750

Size of the auto-converted Parquet files:

384 kB

Number of rows:

27,365