ConGNN: Context-consistent cross-graph neural network for group emotion recognition in the wild

Wang, Yu; Zhou, Shunping; Liu, Yuanyuan; Wang, Kunpeng; Fang, Fang; Qian, Huanyan

doi:10.1016/j.ins.2022.08.003

Cited by 7 publications

(1 citation statement)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, long short-term memory (LSTM) was used to aggregate the features of scenes and faces [18][19][20]. Graph Neural Networks were also employed to fuse different emotional cues and exploit the underlying relations and interactions between the emotional cues [21]. Although the GER methods equipped with these fusion techniques have achieved promising experimental results, their performance still suffers from the limited number of labeled samples.…”

Section: Group Emotion Recognitionmentioning

confidence: 99%

Semi-Supervised Group Emotion Recognition Based on Contrastive Learning

et al. 2022

View full text Add to dashboard Cite

The performance of all learning-based group emotion recognition (GER) methods depends on the number of labeled samples. Although there are lots of group emotion images available on the Internet, labeling them manually is a labor-intensive and cost-expensive process. For this reason, datasets for GER are usually small in size, which limits the performance of GER. Considering labeling manually is challenging, using limited labeled images and a large number of unlabeled images in the network training is a potential way to improve the performance of GER. In this work, we propose a semi-supervised group emotion recognition framework based on contrastive learning to learn efficient features from both labeled and unlabeled images. In the proposed method, the unlabeled images are used to pretrain the backbone by a contrastive learning method, and the labeled images are used to fine-tune the network. The unlabeled images are then given pseudo-labels by the fine-tuned network and used for further training. In order to alleviate the uncertainty of the given pseudo-labels, we propose a Weight Cross-Entropy Loss (WCE-Loss) to suppress the influence of the samples with unreliable pseudo-labels in the training process. Experiment results on three prominent benchmark datasets for GER show the effectiveness of the proposed framework and its superiority compared with other competitive state-of-the-art methods.

show abstract