As an image enhancement technology, multi-modal image fusion primarily aims to retain salient information from multi-source image pairs in a single image, generating imaging information that contains complementary features and can facilitate downstream visual tasks. However, dual-stream methods with convolutional neural networks (CNNs) as backbone networks predominantly have limited receptive fields, whereas methods with Transformers are time-consuming, and both lack the exploration of cross-domain information. This study proposes an innovative image fusion model designed for multi-modal images, encompassing pairs of infrared and visible images and multi-source medical images. Our model leverages the strengths of both Transformers and CNNs to model various feature types effectively, addressing both short- and long-range learning as well as the extraction of low- and high-frequency features. First, our shared encoder is constructed based on Transformers for long-range learning, including an intra-modal feature extraction block, an inter-modal feature extraction block, and a novel feature alignment block that handles slight misalignments. Our private encoder for extracting low- and high-frequency features employs a dual-stream architecture based on CNNs, which includes a dual-domain selection mechanism and an invertible neural network. Second, we develop a cross-attention-based Swin Transformer block to explore cross-domain information. In particular, we introduce a weight transformation that is embedded into the Transformer block to enhance the efficiency. Third, a unified loss function incorporating a dynamic weighting factor is formulated to capture the inherent commonalities of multi-modal images. A comprehensive qualitative and quantitative analysis of image fusion and object detection experimental results demonstrates that the proposed method effectively preserves thermal targets and background texture details, surpassing state-of-the-art alternatives in terms of achieving high-quality image fusion and improving the performance in subsequent visual tasks.

Legal judgment prediction (LJP) is used to predict judgment results based on the description of individual legal cases. In order to be more suitable for actual application scenarios in which the case has cited multiple articles and has multiple charges, we formulate legal judgment prediction as a multiple label learning problem and present a deep learning model that can effectively encode the content of each legal case via a multi-residual convolution neural network and the semantics of law articles via an article encoder. An article-wise attention mechanism is proposed to fuse the two types of encoded information. Experimental results derived on the CAIL2018 datasets show that our model provides a significant performance improvement over the existing neural models in predicting relevant law articles and charges.

Legal Judgment Prediction aims to automatically predict judgment outcomes based on descriptions of legal cases and established law articles, and has received increasing attention. In the preliminary work, several problems still have not been adequately solved. One is how to utilize limited but valuable label information. Existing methods mostly ignore the gap between the description of established articles and cases, but directly integrate them. Second, most studies ignore the mutual constraint among the subtasks, such as logically or semantically, each charge is only related to some specific articles. To address these issues, we first construct a crime similarity graph and then perform a distillation operation to collect discriminate keywords for each charge. Furthermore, we fuse these discriminative keywords instead of established article descriptions into case embedding with a cross-attention mechanism to obtain deep semantic representations of cases incorporating label information. Finally, under a constraint among subtasks, we optimize the one-hot representation of ground-truth labels to guarantee consistent results across the subtasks based on the label-enhancement algorithm. To verify the effectiveness and robustness of our framework, we conduct extensive experiments on two public datasets. The experimental results show that the proposed method outperforms the state-of-art models by 3.89%/7.92% and 1.23%/2.50% in the average MF1-score of the subtasks on CAIL-Small/Big, respectively.