If you have any question regarding this challenge, please send an email to firstname.lastname@example.org
The answers to some frequent asked questions are as follows:
1. Must I participate in all tasks in the challenge?
No, researchers may choose to participate in any one or multiple tasks.
2.Must I submit a paper to the Interspeech special session?
No, you are not required to submit to the special session in order to participate. However, submission to the Interspeech 2020 special session is strongly encouraged, but not mandatory.
3.My team wishes to submit a paper to the Interspeech special session. What should we include?
Papers submitted to the special session should include preliminary results on the development and testing sets; these results may be updated during the camera-ready paper revision period. If they choose to, papers may also report results on other corpora. Papers should not repeat descriptions of the tasks, metrics, datasets, or baselines, but instead cite the challenge paper. For more details, please consult the paper submission instructions of Interspeech 2020.
4.Are there any limitations about the training data?
Participants can use the provided challenge data as well as any database shared in the openslr.org before Feb 1st 2020 for their system training. Please also note that clear descriptions of the data used are required in the final system descriptions document.
5.My team previously has acquired access to the HIMIA dataset. Can we use this data for training or development?
Yes, we allow the participants to include the HIMIA database in their training.
6.I cannot download the data by website link . How can I get the data?
First, you must make sure you have signed and uploaded the data license. Then, please use the registered email account to send an email to email@example.com and we will contact you in 24 hours.
7.Is there a registration deadline?
Participants may register at any time before March 15th.
8.Are we allowed to use Voxceleb for training? Technically, Voxceleb is not hosted at openslr.org (only some metadata is hosted there).
Any publicly open and freely accessible speech database shared on openslr.org before Feb 1st 2020 (including HI-MIA) can be used in this challenge. Voxceleb data is allowed in the training process.
9.Can I use http://openslr.org/87/ data?.
“Any publicly open and freely accessible speech database shared on openslr.org before Feb 1st, 2020(including SLR-85) can be used in this challenge” it was written in the evaluation plan; which means the databases published after SLR-85 cannot be used in this challenge. In this way, you cannot use the SLR-87 as training data.
10.What’s the meaning of “caution.list” in “doc/“ folder?
In caution. list, these audios include some silence regions or are not 100% matching their text transcripts. And the meaning of other files in “doc/” folder was also described in “doc/readme.txt”, please kindly find it and read it carefully.
11.Whether there are any limitations on the number of teams or team members from the same organization?
The teams from the same organization but different departments to participate in the challenge is allowed. However, overlapped members of participated teams are not allowed. System descriptions from the same organization will be reviewed carefully and narrowly.
12.Can I use the adaptive score normalization in the competition?
Yes, the score normalization is standard steps in producing calibrated speaker verification scores, it is allowed to be used in this competition. However, it must be noticed that the "evaluating on each testing sample cannot utilize the information from other testing samples " was published in evaluation plan, so the other trial data can’t be used in each trial.