This paper proposes a multi-scale adaptive fusion sarcasm detection model (MSAF-SDM) to address the challenges of information complexity and insufficient inter-modal collaboration in multimodal sarcasm detection. The model integrates multi-level features from text, audio, and video modalities, leveraging a dynamic attention mechanism and an adaptive weight allocation strategy to capture sarcasm-related cues across modalities. To enhance feature extraction capabilities, the text modality employs a dual multi-scale dilated window attention mechanism, the audio modality utilizes multi-scale temporal convolution, and the video modality incorporates multi-scale spatiotemporal convolution reinforced by auxiliary modal features. Experimental results demonstrate that MSAF-SDM achieves an accuracy of 89.04% and an F1-score of 87.68% on public datasets, significantly outperforming existing state-of-the-art models. Ablation studies further validate the effectiveness of the multimodal feature extraction and adaptive fusion mechanisms. This research provides a novel approach for tackling multimodal sarcasm detection tasks.

Sarcasm often arises from subtle contrasts between literal meaning and speaker intention. As online communication increasingly includes voice-based content, detecting sarcasm across speech and text becomes more important—and more complex. The existing methods usually focus on generic multimodal fusion but often miss how sarcasm manifests differently in each modality. We propose a model that explicitly encodes audio signals into the textual representation space, allowing prosodic cues to inform language understanding. To extract relevant features at different levels, we use a multi-scale convolutional architecture. The experiments show consistent gains over prior models on both text and speech sarcasm detection tasks.

We present HAFF-UNet, a hybrid attention-based semantic segmentation model for accurate water body extraction from remote sensing imagery. By integrating a hybrid attention feature fusion module and multi-scale convolutional attention via atrous spatial pyramid pooling (ASPP), the model enhances spatial detail and global context representation. Evaluated on Gaofen-1 wide field view (WFV) imagery from the Yellow River Basin, HAFF-UNet achieved 94.09% precision, 94.25% recall, 94.16% F1-score, and 89.44% MIoU—outperforming U-Net across all metrics. To further assess generalization, the model was also tested on Landsat 8 imagery from the Huai River Basin, where it achieved 92.97% F1-score and 87.65% MIoU, confirming robustness across different sensors and regions. The results demonstrate its effectiveness in delineating surface water features across large geographic areas with high precision.

This paper proposes a Multi-Scale Adaptive Fusion Sarcasm Detection Model (MSAF-SDM) to address the challenges of information complexity and insufficient inter-modal collaboration in multimodal sarcasm detection. The model integrates multi-level features from text, audio, and video modalities, leveraging a dynamic attention mechanism and an adaptive weight allocation strategy to capture sarcasm-related cues across modalities. To enhance feature extraction capabilities, the text modality employs a dual multi-scale dilated window attention mechanism, the audio modality utilizes multi-scale temporal convolution, and the video modality incorporates multi-scale spatiotemporal convolution reinforced by auxiliary modal features. Experimental results demonstrate that MSAF-SDM achieves an accuracy of 89.04% and an F1-score of 87.68% on public datasets, significantly outperforming existing state-of-the-art models. Ablation studies further validate the effectiveness of the multimodal feature extraction and adaptive fusion mechanisms. This research provides a novel approach for tackling multimodal sarcasm detection tasks.