mirror of
https://github.com/deepinsight/insightface.git
synced 2026-05-15 21:23:52 +00:00
Update README.md
This commit is contained in:
@@ -9,7 +9,15 @@ Submitting features is also allowed, you can send an e-mail to us to request the
|
||||
|
||||
## How to Participate
|
||||
|
||||
Send an e-mail to [..] after preparing your black-box feature extractor or image binary features. With your name, organization and submission comments.
|
||||
Send an e-mail to [..] after preparing your black-box feature extractor or image binary features, with your name, organization and submission comments.
|
||||
|
||||
There are some ways to submit:
|
||||
|
||||
1. (Recommended) Submit black-box face feature extracting tool.
|
||||
2. (Simplest) Submit your recognition model.
|
||||
* Submit MXNet ArcFace model with the same face alignment. In this case, you can just submit the single model file.
|
||||
* In other case, such as PyTorch/TF models or ArcFace models with different face alignment method, please give us an example on how to generate feature embeddings. (eg. provide a function `get_feature(image, landmark)`)
|
||||
3. (Safe) Submit face features. We will give `N` test images named from `0.jpg` to `(N-1).jpg`. Then you should submit a binary feature file with shape `(N,K)` where `K` is your feature dimension. L2 distance will be used on these features.
|
||||
|
||||
## Leaderboard
|
||||
|
||||
|
||||
Reference in New Issue
Block a user