Segment Anything Models (SAMs), as vision foundation models, have demonstrated remarkable performance across various image analysis tasks. Despite their strong generalization capabilities, SAMs encounter challenges in fine-grained detail segmentation for high-resolution class-independent segmentation (HRCS), due to the limitations in the direct processing of high-resolution inputs and low-resolution mask predictions, and the reliance on accurate manual prompts. To address these limitations, we propose MGD-SAM2, which integrates SAM2 with multi-view feature interaction between a global image and local patches to achieve precise segmentation. MGD-SAM2 incorporates the pre-trained SAM2 with four novel modules: the Multi-view Perception Adapter (MPAdapter), the Multi-view Complementary Enhancement Module (MCEM), the Hierarchical Multi-view Interaction Module (HMIM), and the Detail Refinement Module (DRM). Specifically, we first introduce MPAdapter to adapt the SAM2 encoder for enhanced local-global perception required by HRCS images. Then, MCEM and HMIM are proposed to further enrich the local texture and global semantics by aggregating multi-view features within and across multi-scales. Finally, DRM is designed to generate gradually restored high-resolution mask predictions by jointly leveraging mask features and original images, compensating for the detail loss caused by directly upsampling low-resolution prediction maps. Experimental results demonstrate the superior performance and strong generalization of our model on multiple high-resolution and normal-resolution datasets, while requiring only 8.8% trainable parameters. Code is available at https://github.com/sevenshr/MGD-SAM2.

It has been observed that even a small amount of noise introduced into the dataset can significantly degrade the performance of KAN. In this brief note, we aim to quantitatively evaluate the performance when noise is added to the dataset. We propose an oversampling technique combined with denoising to alleviate the impact of noise. Specifically, we employ kernel filtering based on diffusion maps for pre-filtering the noisy data for training KAN network. Our experiments show that while adding i.i.d. noise with any fixed SNR, when we increase the amount of training data by a factor of r, the test-loss (RMSE) of KANs will exhibit a performance trend like test-loss $\sim {\mathcal{O}}\left({{r^{ - \frac{1}{2}}}}\right)$ and as r → +∞. We conclude that applying both oversampling filtering strategies can reduce the detrimental effects of noise. Nevertheless, determining the optimal variance for the kernel filtering process is challenging, and enhancing the volume of training data substantially increases the associated costs, because the training dataset needs to be expanded multiple times in comparison to the initial clean data. As a result, the noise present in the data ultimately diminishes the effectiveness of Kolmogorov-Arnold networks.

The Few-Shot Segmentation (FSS) aims to accomplish the novel class segmentation task with a few annotated images. Current FSS research based on meta-learning focuses on designing a complex interaction mechanism between the query and support feature. However, unlike humans who can rapidly learn new things from limited samples, the existing approach relies solely on fixed feature matching to tackle new tasks, lacking adaptability. In this paper, we propose a novel framework based on the adapter mechanism, namely Adaptive FSS, which can efficiently adapt the existing FSS model to the novel classes. In detail, we design the Prototype Adaptive Module (PAM), which utilizes accurate category information provided by the support set to derive class prototypes, enhancing class-specific information in the multi-stage representation. In addition, our approach is compatible with diverse FSS methods with different backbones by simply inserting PAM between the layers of the encoder. Experiments demonstrate that our method effectively improves the performance of the FSS models (e.g., MSANet, HDMNet, FPTrans, and DCAMA) and achieves new state-of-the-art (SOTA) results (i.e., 72.4% and 79.1% mIoU on PASCAL-5i 1-shot and 5-shot settings, 52.7% and 60.0% mIoU on COCO-20i 1-shot and 5-shot settings). Our code is available at https://github.com/jingw193/AdaptiveFSS.

Recent works have shown that the computational efficiency of 3D medical image (e.g. CT and MRI) segmentation can be impressively improved by dynamic inference based on slice-wise complexity. As a pioneering work, a dynamic architecture network for medical volumetric segmentation (i.e. Med-DANet [44]) has achieved a favorable accuracy and efficiency trade-off by dynamically selecting a suitable 2D candidate model from the pre-defined model bank for different slices. However, the issues of incomplete data analysis, high training costs, and the two-stage pipeline in Med-DANet require further improvement. To this end, this paper further explores a unified formulation of the dynamic inference framework from the perspective of both the data itself and the model structure. For each slice of the input volume, our proposed method dynamically selects an important foreground region for segmentation based on the policy generated by our Decision Network and Crop Position Network. Besides, we propose to insert a stage-wise quantization selector to the employed segmentation model (e.g. U-Net) for dynamic architecture adapting. Extensive experiments on BraTS 2019 and 2020 show that our method achieves comparable or better performance than previous state-of-the-art methods with much less model complexity. Compared with previous methods Med-DANet and TransBTS with dynamic and static architecture respectively, our framework improves the model efficiency by up to nearly 4.1 and 17.3 times with comparable segmentation results on BraTS 2019. Code will be available at https://github.com/Rubics-Xuan/Med-DANet.

High dynamic and interfered environments can seriously deteriorate the synchronization acquisition of the fast frequency hopping (FFH) spread spectrum system, which is particularly detrimental in the anti-interference systems. With the rapid development of the digital processing technologies, some synchronization acquisition algorithms in high dynamic environments have been proposed. However, these algorithms suffer from performance deterioration in the FFH system and do not analyze the acquisition performance under typical interferences. In this paper, a frequency hopping pulses (FHP)-differential adaptive anti-interference combining (DAAC) algorithm-based synchronization acquisition scheme is proposed. In particular, this algorithm combines the interference suppression method and the synchronization acquisition method, and the combining coefficients of the decision metric can be adaptively adjusted based on the interference type and the interference power. In addition, the adaptive anti-interference acquisition thresholds are derived through the constant false-alarm rate (CFAR) criterion. Theoretical and simulation results demonstrate that the proposed FHP-DAAC algorithm enhances the anti-interference and anti-high dynamic performances of the FFH receiver.

Recently, the Synthetic Aperture Radar(SAR) Tomography (TomoSAR) has been used in monitoring the snowpack from X-band to Ku-band. This technique provides unique access to the structure of the imaged scene, and in the case of snowpack, it enables the separation of multiple snow layers as well as the detection of and compensation for soil and vegetation layers. The addition of polarimetric capabilities brings in the ability to detect spatially varying shapes, sizes, and permittivities, to decompose the backscattered signal into volumetric and surface scattering components, and to distinguish between snow, soil, and vegetation. There are a few ground-based field experiments that demonstrate the focused image recover the layering structure of the snowpack with different densities. To better understand the measurement, this paper aims to provide physical-based forward modeling to reconstruct the TomoSAR images with realistic snow profiles. Without loss of generality, we perform the analysis on a two-layer snowpack.

Previous experiments showed that the frequency dependence of snow volume scattering from 18 to 90GHz is power of 2.8, which is much weaker than the power of 4 of Rayleigh scattering. Recently monitoring snow using existing satellites (such as Sentinel-1, COSMO-SkyMed and QuickScat) have been widely studied. The radar signatures from C to Ku band are important for active remote sensing of snow. In this paper, we study the snow aggregation effects and its frequency dependence of volume scattering from C to Ku band (4-18GHz) with the bicontinuous model. The bicontinuous media model is applied to model the snow microstructure with aggregates. The integral equation of snow scattering volume is derived based on the Born approximation which then leads to the scattering coefficients. Scattering coefficients of snow are computed from C to Ku band giving a frequency dependence of ~2.6 power, weaker than that of Rayleigh scattering because of the aggregation effects. In addition, results are also in good agreement with the full wave numerical solutions from X to Ku band.