# 4DMR IJCAI Workshop challenge 2025 — Data

来源：https://www.kaggle.com/competitions/4-dmr-ijcai-workshop-challenge-2025/data

本地 Meta Kaggle 快照；不是本次实时页面。

# Dataset Description

The data is stored as a private kaggle dataset, please sign the licence agreement on the [IJCAI challenge page](https://jingjingchen-pro.github.io/4DMR2025/) to gain access. 
Return the signed licence agreement to our data chair : Mengting Wei Mengting.Wei{at}oulu.fi

The dataset is composed of 4D sequences taken fom high-resolutions scans.
Each sequence is composed of multiple 3D meshes, material file and material source ; in .obj, .mtl and .jpg format respectively.
The dataset contains 24 subject from different cultures, each of them containing a different number of sequences.
Sequences are cut to contain only the micro-expression(s).

The train.csv file will contain the paths of the sequences and their label in the following format :

    path_new, Negative, Positive, Repression, Surprise ,Others, Id
    train\seq0,0,1,0,1,1,1
    train\seq1,1,0,0,1,0,2
    .
    .
    .
    train\seq99,0,1,1,0,0,99


Each participant's submission will be tested on a subset of the test data, the final results will be tested on the full test set.

You can find a sample submission in the sample_submission.csv file.