The increasing scale of graph datasets has significantly improved the performance of graph representation learning methods, but it has also introduced substantial training challenges. Graph dataset condensation techniques have emerged to compress large datasets into smaller yet information-rich datasets, while maintaining similar test performance. However, these methods strictly require downstream applications to match the original dataset and task, which often fails in cross-task and cross-domain scenarios. To address these challenges, we propose a novel causal-invariance-based and transferable graph dataset condensation method, named TGCC, providing effective and transferable condensed datasets. Specifically, to preserve domain-invariant knowledge, we first extract domain causal-invariant features from the spatial domain of the graph using causal interventions. Then, to fully capture the structural and feature information of the original graph, we perform enhanced condensation operations. Finally, through spectral-domain Enhanced contrastive learning, we inject the causal-invariant features into the condensed graph, ensuring that the compressed graph retains the causal information of the original graph. Experimental results on five public datasets and our novel FinReport dataset demonstrate that TGCC achieves up to a 13.41% improvement in cross-task and cross-domain complex scenarios compared to existing methods, and achieves state-of-the-art performance on 5 out of 6 datasets in the single dataset and task scenario.

The exploration of electroencephalogram (EEG) microstates and functional connectivity shows promising potential for both predicting and investigating the neural mechanisms of depression. However, the performance of depression evaluation based on physiological metrics remains unsatisfactory. In this study, we propose a Specific-General Functional Graph Convolutional Network (SGFGCN) to explore biomarkers related to the functional connectivity properties of different EEG microstates. Five microstate topographies, labeled as microstate class A to E, are obtained to describe depressive EEG dynamics, which is highly consistent with the findings of previous studies. Then, incorporating microstate class A to E, our SGFGCN model constructs specific adaptive functional connectivity and general dynamic functional connectivity for each microstate class sequence. It also extracts specific features and general features for each sample using graph convolutional network (GCN). The experimental results, which fuse specific and general depression-related features from resting-state EEG data in the MODMA dataset, demonstrate the superiority of the SGFGCN in identifying depression across various EEG microstate classes compared to previous models. Additionally, based on the recognition results and statistical analysis of the microstates, we confirm that there is a strong correlation between microstate C and depression. These findings suggest the viability of assessing depressive disorders using functional connectivity under certain microstates, offering fresh insights for exploring depression biomarkers.

Electroencephalogram (EEG)-based emotion recognition is a crucial component of affective computing. While the Transformer has proven successful in computer vision, its application in the EEG remains a challenge. This paper introduces a novel model, the Temporal Convolutional Vision Transformer (TC-ViT), seamlessly integrating the architectures of the Temporal Convolutional Network (TCN) and Vision Transformer (ViT), which employs the TCN module to extract preliminary spectral-spatial features from EEG and the ViT module to enhance the extracted features discriminative capability. Emotion classification is executed using a multilayer perceptron. The proposed TC-ViT model demonstrates exceptional performance on the SEED (99.31%) and SEED-IV (93.72%) public datasets, outperforming state-of-the-art methods in EEG-based emotion recognition.

Due to the increasing demand for domestic digital audio and video services, broadcast TV intelligent terminals are rapidly iterating, and global content producers are turning to researching 4K HDR technology. China needs to realize the importance of digital rights, which is more conducive to the development of ultra-high-definition digital services in China. Therefore, this paper will introduce the architecture of China digital rights management (ChinaDRM) and its application in TS encapsulation format. In the meantime, we proposed an analytical method about the TS bitstream which had been encrypted. In the TS bitstream, it also carried the ChinaDRM descriptor and Content Encryption Information (CEI). Finally, we designed a bitstream parsing system to parse the data of CEI and ChinaDRM descriptor.