“A new breakthrough has been made in the field of dialogue emotion recognition, with researchers proposing the DECANet model. By decoupling emotion dependency relationships and cross modal perception mechanisms, it effectively overcomes the challenges of dynamic emotion evolution and multimodal fusion, significantly improving the accuracy and generalization performance of emotion recognition, and providing strong support for emotion recognition in complex interactive scenarios.”