Yang Long

Retinal vessel segmentation underpins quantitative analysis in ophthalmology and is widely used for screening and diagnosis. In practice, manual annotations for thin and tortuous vessels are error-prone, yet the effect of positional label noise on segmentation quality remains underexplored. We address this gap with a lightweight few-shot U-Net-based framework for annotation correction and noise robust learning. Analyses on DRIVE reveal clear performance degradation as label displacement increases. Cross-dataset validation shows that the proposed method attains an Accuracy of 96.51, an AUC of 98.01, and an F1 of 83.55 on CHASE_DB1 and an Accuracy of 97.54, an AUC of 98.45, and an F1 of 83.11 on STARE, achieving competitive performance against state-of-the-art methods. These results quantify the sensitivity of vessel segmentation to positional annotation errors and demonstrate practical robustness under noisy labels.

The proliferation of hateful content in online videos poses severe threats to individual well-being and societal harmony. However, existing solutions for video hate detection either rely heavily on large-scale human annotations or lack fine-grained temporal precision. In this work, we propose LELA, the first training-free Large Language Model (LLM) based framework for hate video localization. Distinct from state-of-the-art models that depend on supervised pipelines, LELA leverages LLMs and modality-specific captioning to detect and temporally localize hateful content in a training-free manner. Our method decomposes a video into five modalities, including image, speech, OCR, music, and video context, and uses a multi-stage prompting scheme to compute fine-grained hateful scores for each frame. We further introduce a composition matching mechanism to enhance cross-modal reasoning. Experiments on two challenging benchmarks, HateMM and MultiHateClip, demonstrate that LELA outperforms all existing training-free baselines by a large margin. We also provide extensive ablations and qualitative visualizations, establishing LELA as a strong foundation for scalable and interpretable hate video localization.

Multi-focus image fusion (MFIF) addresses the challenge of partial focus by integrating multiple source images taken at different focal depths. Unlike most existing methods that rely on complex loss functions or large-scale synthetic datasets, this study approaches MFIF from a novel perspective: optimizing the input space. The core idea is to construct a high-quality MFIF input space in a cost-effective manner by using intermediate features from well-trained, non-MFIF networks. To this end, we propose a cascaded framework comprising two feature extractors, a Feature Distillation and Fusion Module (FDFM), and a focus segmentation network YUNet. Based on our observation that discrepancy and edge features are essential for MFIF, we select a image deblurring network and a salient object detection network as feature extractors. To transform these extracted features into an MFIF-suitable input space, we propose FDFM as a training-free feature adapter. To make FDFM compatible with high-dimensional feature maps, we extend the manifold theory from the edge-preserving field and design a novel isometric domain transformation. Extensive experiments on six benchmark datasets show that (i) our model consistently outperforms 13 state-of-the-art methods in both qualitative and quantitative evaluations, and (ii) the constructed input space can directly enhance the performance of many MFIF models without additional requirements.