Rules and requirements
The goal of the challenge is to foster research on machine learning for 3D audio. All participants should adhere to the following rules to be eligible for the challenge:
- All participants must submit the obtained results for at least one of the 2 tasks, but for both sub-tracks. The results should be accompanied by a paper describing the proposed method.
- Each individual participant could not be included in multiple participating teams. Therefore, a participant is allowed to submit only one set of results.
- Winners will be selected according to the best performance for each single task, separately. Therefore, one winner for each task will be selected.
- There are no restrictions on the methodologies. However, in case of a tie, the Challenge Committee will take into account the novelty and originality of the proposed approach, and the generalization of the method, i.e., its ability to be used for both the 1-mic and 2-mic configuration.
- Participants are not restricted to use the L3DAS21 dataset only. It is in fact allowed to augment this dataset and/or to integrate additional data to train/pre-train the models.
- Results and paper must be submitted within the deadlines shown in the timeline at the challenge website via the IEEE MLSP 2021 submission site.
- Results should be prepared and formatted according to the guidelines detailed in the challenge submission page.
- The accompanying paper must describe the proposed method and must contain all the details to ensure reproducibility. The paper must also include information about the computational complexity of the model (e.g., in terms of number of parameters or execution time on a specific device).
- Papers should be prepared according to the guidelines of the workshop.
- Submitted papers will undergo the standard peer-review process of IEEE MLSP 2021.
- Only models described in papers accepted for IEEE MLSP 2021 will be eligible for winning the challenge.
- Accepted papers will be presented at a special session of the IEEE MLSP 2021 on “Machine Learning for 3D Audio Signal Processing”. Authors who are not interested to participate to the challenge but want to contribute to the topic are encouraged to submit a paper to this track, even without specifically use the proposed datasets.
Any information on the challenge can be requested by email (firstname.lastname@example.org).