Search

Different parts of an image have a strong or weak guiding effect on emotions. The key to emotion recognition of images is to fully exploit the regions associated with emotions. Therefore, this paper proposes a visual sentiment classification model with two branches based on visual transformer, termed as Enhanced Effective Region and Context-Aware Vision Transformers (EERCA-ViT). This model includes a primary branch and an auxiliary branch. The primary branch simulates interdependencies between patches by squeezing and stimulating patches (P-SE), thereby highlighting effective region features in the global feature. The auxiliary branch removes feature patches that have been tagged by the primary branch through the context-aware module (CAM), forcing the network to discover different sentiment discriminative regions. At the same time, a two-part loss function is constructed to improve the robustness of the model. Finally, extensive experiments on six benchmark datasets show that the proposed method outperforms the state-of-the-art image sentiment analysis methods. Furthermore, the effectiveness of the different modules of the framework (P-SE and CAM) has been well demonstrated through extensive comparative experiments. • Removing focused regions can guide the network to capture effective contextual sentiment features. • Fully mining the features of sentiment-related regions is one of the key approaches to sentiment classification. • The method's effectiveness and generalisability were confirmed through extensive experiments on eight datasets. [ABSTRACT FROM AUTHOR]

Abstract This paper focuses on the issue of recognition of facial emotion expressions in video sequences and proposes an integrated framework of two networks: a local network, and a global network, which are based on local enhanced motion history image (LEMHI) and CNN-LSTM cascaded networks respectively. In the local network, frames from unrecognized video are aggregated into a single frame by a novel method, LEMHI. This approach improves MHI by using detected human facial landmarks as attention areas to boost local value in difference image calculation, so that the action of crucial facial unit can be captured effectively. Then this single frame will be fed into a CNN network for prediction. On the other hand, an improved CNN-LSTM model is used as a global feature extractor and classifier for video facial emotion recognition in the global network. Finally, a random search weighted summation strategy is conducted as late-fusion fashion to final predication. Our work also offers an insight into networks and visible feature maps from each layer of CNN to decipher which portions of the face influence the networks' predictions. Experiments on the AFEW, CK+ and MMI datasets using subject-independent validation scheme demonstrate that the integrated framework of two networks achieves a better performance than using individual network separately. Compared with state-of-the-arts methods, the proposed framework demonstrates a superior performance. [ABSTRACT FROM AUTHOR]