Awesome

iMiGUE

Paper

Videos | Paper| Evaluation Protocal

iMiGUE: An Identity-free Video Dataset for Micro-Gesture Understanding and Emotion Analysis <br> Xin Liu<sup>1</sup>, Henglin Shi<sup>2</sup>, Haoyu Chen<sup>2</sup>, Zitong Yu<sup>2</sup>, Xiaobai Li<sup>2</sup>, Guoying Zhao<sup>2</sup> <br> <sup>1</sup> School of Electrical and Information Engineering, Tianjin University, China <br> <sup>2</sup> Center for Machine Vision and Signal Analysis, University of Oulu, Finland

Paper [Information about the paper.]
Abstract
Data [Some details about our dataset and download.]
Method Zoo [Results of different methods on iMiGUE.]
Citation

Abstract

We introduce a new dataset for the emotional artificial intelligence research: identity-free video dataset for MicroGesture Understanding and Emotion analysis (iMiGUE). Different from existing public datasets, iMiGUE focuses on nonverbal body gestures without using any identity information, while the predominant researches of emotion analysis concern sensitive biometric data, like face and speech. Most importantly, iMiGUE focuses on micro-gestures, i.e., unintentional behaviors driven by inner feelings, which are different from ordinary scope of gestures from other gesture datasets which are mostly intentionally performed for illustrative purposes. Furthermore, iMiGUE is designed to evaluate the ability of models to analyze the emotional states by integrating information of recognized micro-gesture, rather than just recognizing prototypes in the sequences separately (or isolatedly). This is because the real need for emotion AI is to understand the emotional states behind gestures in a holistic way. Moreover , to counter for the challenge of imbalanced samples distribution of this dataset, an unsupervised learning method is proposed to capture latent representations from the micro-gesture sequences themselves. We systematically investigate representative methods on this dataset, and comprehensive experimental results reveal several interesting insights from the iMiGUE, e.g., micro-gesture-based analysis can promote emotion understanding. We confirm that the new iMiGUE dataset could advance studies of micro-gesture and emotion AI.

Data

Dataset Statistics and Properties

<div align="center"> <img src="pictures/dataset.png"> </div> <br> We collected 359 videos of post match press conferences of Grand Slam tournaments. Here is the information of the iMiGUE. This dataset contains 72 players from 28 countries and regions covering very continent which enables MGs analysis from diverse cultures. iMiGUE comprises 36 female and 36 male players whose ages are between 17 and 38.

Micro-Gestures examples

The above four pictures are examples of Micro-Gestures, from top to bottom: Folding arms/Covering face/Scratching neck/Rubbing eyes.

Download

Please go to this link.

MethodZoo

Comparison of MG recognition accuracy (%) with state-of-the-art algorithms on the iMiGUE dataset.

Methods	Model+Modality	acc@top1	acc@top5
SVA-E	RNN+Pose	27.38	60.44
LSTM	RNN+Pose	32.36	72.93
BLSTM	RNN+Pose	32.39	71.34
ST-GCN	GCN+Pose	46.97	84.09
2S-GCN	GCN+Pose	47.78	88.43
Shift-GCN	GCN+Pose	51.51	88.18
GCN-NAS	GCN+Pose	53.90	89.21
MS-G3D	GCN+Pose	54.91	89.98
C3D	3DCNN+RGB	20.32	55.31
R3D-101	3DCNN+RGB	25.27	59.39
I3D	3DCNN+RGB	34.96	63.69
TSN	2DCNN+RGB	51.54	85.42
TRN	2DCNN+RGB	55.24	89.17
TSM	2DCNN+RGB	61.10	91.24
P&C	Encoder-Decoder+Pose	31.67	64.93
U-S-VAE Z	Encoder-Decoder+Pose	32.43	64.30

</center>

Comparison of emotion understanding accuracy (%) with methods on iMiGUE dataset

Methods	Model+Modality	Accuracy
TRN	CNN + RGB	0.53
TSM	CNN + RGB	0.53
I3D	CNN + RGB	0.57
ST-GCN	GCN+Pose	0.50
MS-G3D	GCN+Pose	0.55
U-S-VAE + LSTM	RNN + Micro-Gesture	0.55
TSM + LSTM	RNN + Micro-Gesture	0.60