How easy can be used by other participants (e.g.When evaluating a proposal for a new external model/ dataset we will take into consideration the following things: Please reach us via the Google Group or in the Slack Workspace if you want to propose new additional external datasets or pre-trained models. When your proposal is accepted we will update the lists here and notify participants (via Slack Workspace and Google Group) about the new material/model allowed. The proposal period will be open till the 7th March AoE (anywhere on earth). In addition to allowed external datasets and pre-trained models listed in this page below, we encourage participants to propose new ones.

Propose Your External Data and Pre-Trained Models ! Organizers can participate to this Challenge but they are not eligible for prizes and will not appear on the official ranking, their submission will be regarded as a baseline system v2. Note that Organizers will use this mandatory system description paper to assess the correctness of each submission. We require each participant to submit a 2 to 6 pages system description paper (same deadline as the system submission). We do not require participants to open-source their system, however it is highly encouraged. Thus we rely only on your goodwill and correctness to keep this Task fair among all participants.

Organizers can partipicate but are not eligible for prizes and will not appear in the official ranking.Įvaluation data comes from previously published datasets (as in past CHiME-6 Challenge).

We do permit the use of all official data and external data for language model training.

We do not permit the use of pre-trained language models such as BERT.

Automatic device/microphone selection is allowed.

See for an actual example the baseline results on the acoustic robustness sub-track Section 3.1.2 recipe README.md.

only outer array mics on CHiME-6) and a different one for another scenario accounts for domain identification. Note that selecting a subset of microphones on a scenario (e.g. Automatic or manual domain identification is not allowed.Close-talk microphones cannot be used during inference.Other datasets and models can be proposed until March 7th (see Additional Models).Pretrained models specified on this page may be used.External data specified on this page and in the Data page may be used.Please read this page and Data page for detailed rules regarding this task and the associated data.

Slack chime