MSTF-Net: A UAV-Oriented Multi-Spectral Video Segmentation Method via Modality-Robust, Scale-Adaptive, and Consistent Fusion

Multi-spectral video segmentation is essential for robust scene understanding in unmanned aerial vehicle (UAV) applications such as city planning, land use monitoring, traffic monitoring, and crowd estimation. While the fusion of RGB and thermal modalities offers complementary information for perception under varying lighting and visibility conditions, two fundamental challenges remain: (1) the modal fusion dilemma, arising from significant discrepancies between RGB and thermal features that obscure complementary cues, and (2) temporal variation, induced by rapid motion and viewpoint changes on UAV platforms, which leads to appearance inconsistency and misalignment across frames. To address these issues, this study proposed MSTF-Net, a modality-robust scale-adaptive fusion framework for multi-spectral video segmentation that effectively models cross-modal fusion and temporal consistency. The Modality Spatial Complementary Suppression and Enhancement (MSCSE) module generates unified instance queries via cross-modal attention and suppresses modality-specific noise using residual-guided discrepancy filtering and consistency constraints. To model temporal dynamics, the Multi-scale Temporal Cross-modality Semantic Consistency (MTCSC) module adaptively adjusts the temporal receptive field based on frame distance, capturing both coarse global context and fine local structure across time. Extensive ablation experiments on public RGB-T datasets demonstrate that MSTF-Net achieves state-of-the-art segmentation performance, especially under challenging conditions such as small targets, occlusion, and modality degradation. Specifically, reached 56.42\% mIoU on the MVSeg dataset and 51.80\% mIoU on the CART dataset, respectively.

Zero-Shot Size Transfer for Neural ODEs on Sparse Random Graphs: Graphon Limits and Adjoint Convergence

Graph Neural Differential Equations (GNDEs) model continuous-time graph dynamics by parameterizing Neural ODE velocity fields with Graph Neural Networks. Their local, size-independent filters suggest a zero-shot size-transfer principle: train on a small graph and deploy on larger, similar graphs without retraining. We develop a quantitative theory for this principle on sparse random graphs sampled from graphons. We consider Graphon Neural Differential Equations (Graphon-NDEs) and adjoint Graphon-NDEs as the infinite-node limits of the forward and adjoint GNDE systems, and establish well-posedness. For an $n$-node random graph with sparsity parameter $α_n$, we prove trajectory-wise convergence of GNDE solutions to Graphon-NDE solutions at rate $O((α_n n)^{-1/2})$, up to logarithmic factors, with high probability. We also establish uniform-in-time convergence bounds for adjoint systems governing hidden-state and parameter gradients. We further study discretize-then-optimize (DTO) and optimize-then-discretize (OTD) training. Under explicit Euler discretization with $M$ steps, we show that DTO and OTD are asymptotically consistent, with hidden-state and local parameter-gradient discrepancies of orders $O(1/M)$ and $O(1/M^2)$, respectively, up to sparsity and logarithmic factors. Experiments on HSBM and tent graphons support the theoretical rates, while zero-shot transfer experiments across four graphon classes demonstrate accurate deployment of learned GNDEs on larger independently sampled graphs.

A Novel Hybrid Approach for Retinal Vessel Segmentation with Dynamic Long-Range Dependency and Multi-Scale Retinal Edge Fusion Enhancement

Accurate retinal vessel segmentation provides essential structural information for ophthalmic image analysis. However, existing methods struggle with challenges such as multi-scale vessel variability, complex curvatures, and ambiguous boundaries. While Convolutional Neural Networks (CNNs), Transformer-based models and Mamba-based architectures have advanced the field, they often suffer from vascular discontinuities or edge feature ambiguity. To address these limitations, we propose a novel hybrid framework that synergistically integrates CNNs and Mamba for high-precision retinal vessel segmentation. Our approach introduces three key innovations: 1) The proposed High-Resolution Edge Fuse Network is a high-resolution preserving hybrid segmentation framework that combines a multi-scale backbone with the Multi-scale Retina Edge Fusion (MREF) module to enhance edge features, ensuring accurate and robust vessel segmentation. 2) The Dynamic Snake Visual State Space block combines Dynamic Snake Convolution with Mamba to adaptively capture vessel curvature details and long-range dependencies. An improved eight-directional 2D Snake-Selective Scan mechanism and a dynamic weighting strategy enhance the perception of complex vascular topologies. 3) The MREF module enhances boundary precision through multi-scale edge feature aggregation, suppressing noise while emphasizing critical vessel structures across scales. Experiments on three public datasets demonstrate that our method achieves state-of-the-art performance, particularly in maintaining vascular continuity and effectively segmenting vessels in low-contrast regions. This work provides a robust method for clinical applications requiring accurate retinal vessel analysis. The code is available at https://github.com/frank-oy/HREFNet.