Real-world Affective Faces Multi Label
Details
Humans' facial representations are often appear as
combinations, blends, or compounds of different basic emotions.
Real-world Affective Faces Multi Label (RAF-ML) is a multi-label facial expression dataset with around 5K great-diverse facial images downloaded from the Internet with blended emotions and variability in subjects' identity, head poses, lighting conditions and occlusions. During annotation, 315 well-trained annotators are employed to ensure each image can be annotated enough independent times. And images with multi-peak label distribution are selected out to constitute the RAF-ML.
In RAF-ML, we provide 4908 number of real-world images with blended emotions, 6-dimensional expression distribution vector for each image, 5 accurate landmark locations and 37 automatic landmark locations, and baseline classifier outputs for multi-label emotion recognition.
For more details of the dataset, please refer to the paper "Blended Emotion in-the-wild: Multi-label Facial Expression Recognition Using Crowdsourced Annotations and Deep Locality Feature Learning".
We also provide a singl-label compound emotion database, with corresponding AU label in RAF-CE
Sample Images
Terms & Conditions
-
The RAF-ML is available for non-commercial research purposes only.
-
All images of the RAF-ML are obtained from the Internet which are not property of PRIS, Beijing University of Posts and Telecommunications. The PRIS is not responsible for the content nor the meaning of these images.
You agree not to reproduce, duplicate, copy, sell, trade, resell or exploit for any commercial purposes, any portion of the images and any portion of derived data.
-
You agree not to further copy, publish or distribute any portion of the RAF-ML. Except, for internal use at a single site within the same organization it is allowed to make copies of the dataset.
-
The PRIS reserves the right to terminate your access to the RAF-ML at any time.
Download
To facilitate the research on multi-label learning and label distributiuon learning, one can directly download the labels and corresponding feature vectors of RAF-ML.
multilabel.txt: Multiple expression label of each image. Each column of the blended emotion label represents Surprise, Fear, Disgust, Happiness, Sadness and Anger, respectively.
distribution.txt: 6-dimensional expression distribution of each image. Each column of the 6-dimensional expression distribution represents Surprise, Fear, Disgust, Happiness, Sadness and Anger, respectively.
partition_label.txt: The training/test partition we used in our experiments. Label 0 for train; label 1 for test.
LBP.zip: 5900-dimensional LBP feature of each image.
baseDCNN.zip: 2000-dimensional baseDCNN feature of each image.
DBM_CNN.zip: 2000-dimensional DBM-CNN feature of each image.
In order to download the source images for further study, one must apply for RAF-ML database as follows.
How to get the Password
This database is publicly available. It is free for professors and researcher scientists affiliated to a University.
Permission to use but not reproduce or distribute the RAF-ML database is granted to all researchers given that the following steps are properly followed:
Send an e-mail to Shan Li (email1) before downloading the database. You will need a password to access the files of this database. Your Email MUST be set from a valid University account and MUST include the following text:
Subject: (RAF-ML) Application to download the RAF-ML Dataset Name: <your first and last name> Affiliation: <University where you work> Department: <your department> Position: <your job title> Email: <must be the email at the above mentioned institution>
I have read and agree to the terms and conditions specified in the RAF face database webpage. This database will only be used for research purposes. I will not make any part of this database available to a third party. I'll not sell any part of this database or make any profit from its use.
Citation
@article{DBLP:journals/ijcv/ShangD19,
author = {Shan Li and Weihong Deng},
title = {Blended Emotion in-the-Wild: Multi-label Facial Expression Recognition
Using Crowdsourced Annotations and Deep Locality Feature Learning},
journal = {International Journal of Computer Vision},
volume = {127},
number = {6-7},
pages = {884--906},
year = {2019}
}