Forums

The VoxCeleb Speaker Recognition Challenge 2021 - Track 1 (Verification fully supervised, closed) Forum

Go back to competition Back to thread list Post in this thread

> About the usage of video data.

For the fully supervised speaker verification task, can we use the video or face data of VoxCeleb2 dev dataset? If I can use the video or face data, which track does it belongs to? Track1 or track2?

Posted by: ab3007 @ Aug. 16, 2021, 2:38 a.m.

Hi, thank you for posting this.

Unfortunately, you cannot use the video of the VoxCeleb1 because we're focusing on speaker verification only with audio.

Posted by: vgg_oxford @ Aug. 16, 2021, 9:08 a.m.
Post in this thread