Research on Animated GIFs Emotion Recognition Based on ResNet-ConvGRU

Qian Zhang, Ren Qing-Dao-Er-Ji, Na Li

Abstract: Animated Graphics Interchange Format (GIF) images have become an important part of network information interaction, and areone of the main characteristics of analyzing social media emotions. At present, most of the research on GIF affection recognitionfails to make full use of spatial-temporal characteristics of GIF images, which limits the performance of model recognition to acertain extent. A GIF emotion recognition algorithm based on ResNet-ConvGRU is proposed in this paper. First, GIF data ispreprocessed, converting its image sequences to static image format for saving. Then, the spatial features of images and thetemporal features of static image sequences are extracted with ResNet and ConvGRU networks, respectively. At last, the animatedGIFs data features are synthesized and the seven emotional intensities of GIF data are calculated. The GIFGIF dataset is used toverify the experiment. From the experimental results, the proposed animated GIFs emotion recognition model based on ResNet-ConvGRU, compared with the classical emotion recognition algorithms such as VGGNet-ConvGRU, ResNet3D, CNN-LSTM,and C3D, has a stronger feature extraction ability, and sentiment classification performance. This method provides a finer-grainedanalysis for the study of public opinion trends and a new idea for affection recognition of GIF data in social media.