Training data contains ~11500 examples, each of them consists of equivalently recorded 5-frame sequence.
In training data set, spoof class (fraud) has three separate categories:
2dmask
- printed photo face mask;printed
- printed photo display;replay
- video record replay.Testing images are similar.
Spoof deems as a positive class of label=1
. P(falsealarm)
is a probability of a false alarm, i.e. real face classified as a spoof. P(miss)
— the probability of missing a spoof.
Any open data sets and open pre-trained models are allowed for algorithm training.
Training data contains:
Challenge discussion is held in the #idrnd_challenge
channel of ODS.AI slack (registration is available here).
Teams are up to 6 people. Public code sharing is allowed strictly into the slack discussion channel.
Everyone can participate, except for LLC "DSP Labs" and "Singularis Lab" employees.
Challenge takes place from May 9 to June 20 inclusively. Team submission deadline - June 13, 23:59.
The solution must include an archive containing the algorithm (code + necessary files) and startup point description. The root directory of a solution archive must include file meta.json
of the following structure:
{
"image": "<docker image>",
"entrypoint": "<entry point or sh script>"
}
For example:
{
"image": "ksanvatds/idrnd-antispoof",
"entrypoint": "python3 predict.py --path-images-csv $PATH_INPUT/meta.csv --path-test-dir $PATH_INPUT --path-submission-csv $PATH_OUTPUT/solution.csv"
}
The solution runs in a docker-container with ksanvatds/idrnd-antispoof
as a base image, similar to GPU image for Kaggle.
Organizers provide a correct baseline solution. Additional info is available at the GitHub page of a challenge.
CUDA 10, CUDNN 7.4, and recent library versions are provided. If desired, own image could be dispensed.