Brain decoding is a key neuroscience field that reconstructs the visual stimuli from brain activity with fMRI, which helps illuminate how the brain represents the world. fMRI-to-image reconstruction has achieved impressive progress by leveraging diffusion models. However, brain signals infused with prior knowledge and associations exhibit a significant information asymmetry when compared to raw visual features, still posing challenges for decoding fMRI representations under the supervision of images. Consequently, the reconstructed images often lack fine-grained visual fidelity, such as missing attributes and distorted spatial relationships. To tackle this challenge, we propose BrainCognizer, a novel brain decoding model inspired by human visual cognition, which explores multilevel semantics and correlations without fine-tuning of generative models. Specifically, BrainCognizer introduces two modules: the Cognitive Integration Module which incorporates prior human knowledge to extract hierarchical region semantics; and the Cognitive Correlation Module which captures contextual semantic relationships across regions. Incorporating these two modules enhances intra-region semantic consistency and maintains interregion contextual associations, thereby facilitating fine-grained brain decoding. Moreover, we quantitatively interpret our components from a neuroscience perspective and analyze the associations between different visual patterns and brain functions. Extensive quantitative and qualitative experiments demonstrate that BrainCognizer outperforms state-of-the-art approaches on multiple evaluation metrics. Our code is released publicly at https://github.com/Grace160/BrainCognizer.

Although graph neural networks based methods can solve the uneven text length problem of text classification datasets, they are difficult to address the data sparsity problem of short texts. Although some researchers try to reduce the sparsity of the graph by adding labels to its structure, most of them only treat labels as node features other than words and documents, which is not sufficient to construct denser matrices. To address the above problems, three label data augmentation strategies are proposed to build a dense graph, and the attention mechanisms are used to update node features. In addition, a node feature updating method that simultaneously uses global and local weights is proposed. Multiple comparative experiments on five benchmark datasets demonstrate that the method proposed in this article is optimal and the accuracy and micro-F1 have improved by at least 0.012 on four benchmark datasets.

Since multilabel text classification datasets often face the problem of label imbalance, therefore, using either sequence-based deep learning (DL) model or graph neural network (GNN)-based DL model alone will not achieve satisfactory classification results. To solve the above problem, firstly, two coattention networks are constructed to simultaneously obtain the sequence-based and GNN-based eigenvectors. Second, labels are added to the graph as global features, and a graph data augmentation strategy is proposed. When obtaining GNN-based eigenvectors, at first, connection and attention weights are obtained through adjacency matrix and the attention of neighborhoods. Then, node features are updated based on convolution and multihead attention, respectively. Multiple comparison experiments on four benchmark datasets prove that the model constructed in this article achieves the optimal classification results and can solve the label imbalance problem.

Semi-supervised learning methods, compared to fully supervised learning, offer significant potential to alleviate the burden of manual annotations on clinicians. By leveraging unlabeled data, these methods can aid in the development of medical image segmentation systems for improving efficiency. Boundary segmentation is crucial in medical image analysis. However, accurate segmentation of boundary regions is under-explored in existing methods since boundary pixels constitute only a small fraction of the overall image, resulting in suboptimal segmentation performance for boundary regions. In this paper, we introduce boundary-guided contrastive learning for semi-supervised medical image segmentation (BoCLIS). Specifically, we first propose conservative-to-radical teacher networks with an uncertainty-weighted aggregation strategy to generate higher quality pseudo-labels, enabling more efficient utilization of unlabeled data. To further improve the performance of segmentation in boundary regions, we propose a boundary-guided patch sampling strategy to guide the framework in learning discriminative representations for these regions. Lastly, the patch-based contrastive learning is proposed to simultaneously compute the (dis)similarities of the discriminative representations across intra- and inter-images. Extensive experiments on three public datasets show that our method consistently outperforms existing methods, especially in the boundary region, with DSC improvements of 20.47%, 16.75%, and 17.18%, respectively. A comprehensive analysis is further performed to demonstrate the effectiveness of our approach. Our code is released publicly at https://github.com/youngyzzZ/BoCLIS.

Single-stage weakly-supervised semantic segmentation (WSSS) with image-level labels has become a new research hotspot in the community for its lower cost and higher training efficiency. However, the pseudo label of WSSS generally suffers from somewhat noise, which limits the segmentation performance. In this paper, to explore the integral foreground activation, we propose the Channel Suppression (CS) module for preventing only activating the most discriminative regions, thereby improving the initial pseudo labels. To rectify the in-correct prediction, we explore the Self-Attention Prediction Correction (SAPC) module, which adaptively generates the category-wise prediction rectification weights. After extensive experiments, the proposed efficient single-stage framework achieves excellent performance with 67.6% mIoU and 39.9% mIoU on PASCAL VOC 2012 and MS COCO 2014 datasets, significantly exceeding several recent single-stage methods.

Abstract Online education is becoming more and more popular with the development of the Internet. In particular, due to the COVID-19 pandemic, many countries around the world are increasing the popularity of online education, which makes the research on sentiment classification of course reviews of online education websites an important research direction in natural language processing tasks. Traditional sentiment classification models are mostly based on English. Unlike English, Chinese characters are based on pictograms. Radicals of Chinese characters can also express certain semantics, and characters with the same radical often have similar meanings. Therefore, RSCOEWR, a word-level and radical-level based sentiment classification model for course reviews of Chinese online education websites is proposed, which solves the problem of data sparsity of reviews by feature extraction of multiple dimensions. In addition, a deep learning model based on CNN, BILSTM, BIGRU and Attention is constructed to solve the problem of high dimension and assigning the same attention to context of traditional sentiment classification model. Extensive comparative experiment results show that RSCOEWR outperforms the state-of-the-art sentiment classification models, and the experimental results on public Chinese sentiment classification datasets prove the generalization ability of RSCOEWR.

Abstract On-line gambling is the key illegal behaviour of public security department in most countries due to the potential threat to cyberspace security and social stability. Hence, the research on gambling domain names (GDN) classification is quite important and in great demand for academia and industry. Till now, there is very little research work on this topic. Most of the GDN training datasets in previous work were chosen from GDN blacklists provided by publicly available data sources, and the authors did not verify the authenticity and accuracy of these datasets, and the classification results are not particularly satisfactory. In this paper, certificated and textual analysis-based classification method CT-GDNC is proposed to get GDN training data set with an accuracy of 0.9776 and significantly improve the classification results of GDN. The exhaustive comparative experiments on 10K GDN obtained via Bert fine-tuning model and 10K benign data collected from Alex Top 1 million list show that the proposed method achieves new baseline result for GDN classification with classification accuracy 0.9936, precision 0.9936, F1 0.9936 and recall 0.9939.

Weakly supervised object localization (WSOL) aims to cover the entire target object only under the image-level supervision. Most WSOL methods are stuck in mining the CAMs (class activation maps) of deep semantic features for they only focus on limited discriminative regions playing key role in classification. Recently, a new paradigm has emerged by localizing objects using the low-level feature through two stages. Existing two-stages methods usually train a classification network first to yield CAMs as pseudo labels to guide the learning of segment network, yet it does not consider the activations with more background noise or less discriminative area. In this paper, we propose a Pixel Alignment strategy to refine the object localization by improving the shallow-feature based CAMs generator with the joint supervision of pseudo-label mask, classification evaluation, and absolution size constraint on the activation map. More specifically, we utilize the class-specific pixel gradient to achieve a robust activation pseudo mask to background noise, which further supervises the activation generator with confident foreground and background regions. We also adapt a post-processing to excavate the target region in the conflict area (i.e., the non-overlap area of CAMs and the activations). Extensive experiments on CUB-2002011 and ILSVRC datasets indicate that our method outperforms the state-of-the-art among the two-stage works.