From 5221c2797f0e27f0e92893c7b2864f064a3174e3 Mon Sep 17 00:00:00 2001 From: xiongxinlei Date: Wed, 23 Mar 2022 15:01:00 +0800 Subject: [PATCH] add voxceleb dataset and trial info, test=doc --- examples/voxceleb/README.md | 28 ++++++++++++++++++++++++++++ 1 file changed, 28 insertions(+) diff --git a/examples/voxceleb/README.md b/examples/voxceleb/README.md index fc847cd8a..a2e58e00d 100644 --- a/examples/voxceleb/README.md +++ b/examples/voxceleb/README.md @@ -26,3 +26,31 @@ ffmpeg -y -i %s -ac 1 -vn -acodec pcm_s16le -ar 16000 %s You can do the conversion using ffmpeg https://gist.github.com/seungwonpark/4f273739beef2691cd53b5c39629d830). This operation might take several hours and should be only once. 3. Put all the wav files in a folder called `wav`. You should have something like `voxceleb2/wav/id*/*.wav` (e.g, `voxceleb2/wav/id00012/21Uxsk56VDQ/00001.wav`) + + +## voxceleb dataset summary + + +|dataset | vox1 - dev | vox1 - test |vox2 - dev| vox2 - test| +|---------|-----------|------------|-----------|----------| +|spks | 1211 |40 | 5994 | 118| +|utts | 148642 | 4874 | 1092009 |36273| +| time(h) | 340.4 | 11.2 | 2360.2 |79.9 | + + +## trial summary + +| trial | filename | nums | positive | negative | +|--------|-----------|--------|-------|------| +| VoxCeleb1 | veri_test.txt | 37720 | 18860 | 18860 | +| VoxCeleb1(cleaned) | veri_test2.txt | 37611 | 18802 | 18809 | +| VoxCeleb1-H | list_test_hard.txt | 552536 | 276270 | 276266 | +|VoxCeleb1-H(cleaned) |list_test_hard2.txt | 550894 | 275488 | 275406 | +|VoxCeleb1-E | list_test_all.txt | 581480 | 290743 | 290737 | +|VoxCeleb1-E(cleaned) | list_test_all2.txt |579818 |289921 |289897 | + + + + + +