Small or large superpixel graphs? Gaussian influence walk with rebound can assist

Pixel-level annotation of lung cavities (LCs) in computed tomography (CT) images is challenging due to their morphological diversity and complexity. Weakly supervised semantic segmentation (WSSS) methods, which utilize sparse annotations (e.g., image-level labels), offer a promising solution. However, existing WSSS approaches often generate coarse pseudo-labels and lack sufficient spatial supervision, resulting in under- or over-segmentation of irregular lesions. To address these limitations, we introduce several key innovations. First, we propose a novel Graph-based Affinity Network (GA-Net) that, unlike conventional methods relying on low-level pixel features, models long-range contextual relationships and structural dependencies using a superpixel graph and learned edge inference kernel, enabling structure-aware pseudo-label refinement for complex lesion morphology. Second, we introduce region-wise affinity propagation, which refines segmentation by propagating activations within semantically coherent 3D regions, offering more precise control over under-/over-segmentation compared to global affinity methods. Additionally, we incorporate Exponential Moving Average (EMA) ensembling for training stability and a scribble-based segmentation module that utilizes pseudo-label contours to provide direct boundary supervision. Extensive experiments on three benchmark datasets demonstrate that our method outperforms existing state-of-the-art medical WSSS techniques, achieving precise and reliable segmentation of complex LCs in CT scans.

Pixel-level annotation of lung cavities (LCs) in computed tomography (CT) images is challenging due to their morphological diversity and complexity. Weakly supervised semantic segmentation (WSSS) methods, which utilize sparse annotations (e.g., image-level labels), offer a promising solution. However, existing WSSS approaches often generate coarse pseudo-labels and lack sufficient spatial supervision, resulting in under- or over-segmentation of irregular lesions. To address these limitations, we introduce several key innovations. First, we propose a novel Graph-based Affinity Network (GA-Net) that, unlike conventional methods relying on low-level pixel features, models long-range contextual relationships and structural dependencies using a superpixel graph and learned edge inference kernel, enabling structure-aware pseudo-label refinement for complex lesion morphology. Second, we introduce region-wise affinity propagation, which refines segmentation by propagating activations within semantically coherent 3D regions, offering more precise control over under-/over-segmentation compared to global affinity methods. Additionally, we incorporate Exponential Moving Average (EMA) ensembling for training stability and a scribble-based segmentation module that utilizes pseudo-label contours to provide direct boundary supervision. Extensive experiments on three benchmark datasets demonstrate that our method outperforms existing state-of-the-art medical WSSS techniques, achieving precise and reliable segmentation of complex LCs in CT scans.

Convolutional neural networks (CNNs) and graph convolutional networks (GCNs) have been widely applied to hyperspectral image classification tasks, but both exhibit certain limitations. To address these issues, this paper proposes a multi-scale feature fusion architecture (MCGNet). Symmetry serves as the core design principle of MCGNet, where its parallel CNN-GCN branches and multi-scale fusion mechanism strike a balance between local spectral-spatial features and global graph structural dependencies, effectively reducing redundancy and enhancing generalization capabilities. The architecture comprises four modules: the Spectral Noise Suppression (SNS) module enhances the signal-to-noise ratio of spectral features; the Local Spectral Extraction (LSE) module employs deep separable convolutions to extract local spectral-spatial features; Superpixel-level Graph Convolution (SGC), performing graph convolution on superpixel graphs to precisely capture dependencies between object regions; Pixel-level Graph Convolution (PGC), constructed via adaptive sparse pixel graphs based on spectral and spatial similarity to accurately capture irregular boundaries and fine-grained non-local relationships between pixels. These modules form a symmetric, hierarchical feature learning pipeline integrated within a unified framework. Experiments on three public datasets—Indian Pine, Pavia University, and Salinas—demonstrate that MCGNet outperforms baseline methods in overall accuracy, average precision, and Kappa coefficient. This symmetric design not only enhances classification performance but also endows the model with strong theoretical interpretability and cross-dataset robustness, highlighting the significance of symmetry principles in hyperspectral image analysis.

Hyperspectral image classification is crucial in remote sensing but faces challenges in forest ecosystem studies due to high-dimensional data, spectral variability, and spatial heterogeneity. Watershed Superpixel Segmentation and Sparse Graph Convolutional Networks (WSSGCN), a novel framework designed for efficient forest image classification, is introduced in this paper. Watershed superpixel segmentation is first used by the method to divide hyperspectral images into semantically consistent regions, reducing computational complexity while preserving terrain boundary information. On this basis, a dual-branch model is designed: a local branch with multi-scale convolutional neural networks (CNN) extracts spatial–spectral features, while a global branch constructs superpixel graphs and uses GCNs to model the global context. To enhance efficiency, a sparse tensor-based storage method is proposed for the adjacency matrix, reducing complexity from quadratic to linear. Additionally, an attention-based adaptive fusion strategy dynamically balances local and global features. Experiments on multiple datasets show that WSSGCN outperforms mainstream methods in overall accuracy (OA), average accuracy (AA), and Kappa coefficient. Notably, it achieves a 3.5% OA improvement and a 0.04 Kappa coefficient increase compared to SPEFORMER on the WHU-Hi-HongHu dataset. Practicality in resource-limited scenarios is ensured by sparse graph modeling. This work offers an efficient solution for forest monitoring, supporting applications like biodiversity assessment and deforestation tracking, and advances remote sensing-based forest ecosystem analysis. The proposed approach shows strong potential for real-world ecological conservation and forest management.

Subcellular structure segmentation is a fundamental task in biological imaging. Existing self-supervised representation learning combined with classical k-means clustering achieved unsupervised image segmentation, but it was constrained by time-consuming test-time pixel-wise feature extraction and clustering synchronization. This study introduces SCCS, a lightweight graph neural network-based spectral clustering framework for end-to-end subcellular structure segmentation upon superpixel graphs, greatly relieving the computational complexity in test-time numerical spectral clustering and inter-graph label inconsistency. Specifically, SCCS exploits the self-supervised masked autoencoder for representation learning and the construction of superpixel graphs (spG). Unlike per-graph scalar affinity-based spectral clustering, the proposed SCCS parameterizes the mapping from learned deep spG representations to coordinates in the spectral embedding space and the clustering assignments. The SCCS is optimized under unsupervised eigendecomposition and incremental clustering criteria, which synchronize the intra- and inter-graph spectral clustering. The proposed approach is evaluated on a publicly available volumetric electron microscopy dataset. Experiments demonstrate the effectiveness and performance gains of the proposed SCCS over the state-of-the-art in discovering a variety of subcellular structures.

Proposal-free instance segmentation methods have significantly advanced the field of biological image analysis. Recently, the Segment Anything Model (SAM) has shown an extraordinary ability to handle challenging instance boundaries. However, directly applying SAM to biological images that contain instances with complex morphologies and dense distributions fails to yield satisfactory results. In this work, we propose BioSAM, a new biological instance segmentation framework generating SAM prompts from a superpixel graph. Specifically, to avoid over-merging, we first generate sufficient superpixels as graph nodes and construct an initialized graph. We then generate initial prompts from each superpixel and aggregate them through a graph neural network (GNN) by predicting the relationship of superpixels to avoid over-segmentation. We employ the SAM encoder embeddings and the SAM-assisted superpixel similarity as new features for the graph to enhance its discrimination capability. With the graph-based prompt aggregation, we utilize the aggregated prompts in SAM to refine the segmentation and generate more accurate instance boundaries. Comprehensive experiments on four representative biological datasets demonstrate that our proposed method outperforms state-of-the-art methods.

Hyperspectral images (HSI) clustering is an important but challenging task. The state-of-the-art (SOTA) methods usually rely on superpixels, however, they do not fully utilize the spatial and spectral information in HSI 3-D structure, and their optimization targets are not clustering-oriented. In this work, we first use 3-D and 2-D hybrid convolutional neural networks to extract the high-order spatial and spectral features of HSI through pre-training, and then design a superpixel graph contrastive clustering (SPGCC) model to learn discriminative superpixel representations. Reasonable augmented views are crucial for contrastive clustering, and conventional contrastive learning may hurt the cluster structure since different samples are pushed away in the embedding space even if they belong to the same class. In SPGCC, we design two semantic-invariant data augmentations for HSI superpixels: pixel sampling augmentation and model weight augmentation. Then sample-level alignment and clustering-center-level contrast are performed for better intra-class similarity and inter-class dissimilarity of superpixel embeddings. We perform clustering and network optimization alternatively. Experimental results on several HSI datasets verify the advantages of the proposed SPGCC compared to SOTA methods. Our code is available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/jhqi/spgcc</uri>.

Background/Objectives: Accurate detection and classification of nodules in medical images, particularly rheumatoid nodules, are critical due to the varying nature of these nodules, where their specific type is often unknown before analysis. This study addresses the challenges of multi-class prediction in nodule detection, with a specific focus on rheumatoid nodules, by employing a comprehensive approach to feature extraction and classification. We utilized a diverse dataset of nodules, including rheumatoid nodules sourced from the DermNet dataset and local rheumatologists. Method: This study integrates 62 features, combining traditional image characteristics with advanced graph-based features derived from a superpixel graph constructed through Delaunay triangulation. The key steps include image preprocessing with anisotropic diffusion and Retinex enhancement, superpixel segmentation using SLIC, and graph-based feature extraction. Texture analysis was performed using Gray-Level Co-occurrence Matrix (GLCM) metrics, while shape analysis was conducted with Fourier descriptors. Vascular pattern recognition, crucial for identifying rheumatoid nodules, was enhanced using the Frangi filter. A Hybrid CNN–Transformer model was employed for feature fusion, and feature selection and hyperparameter tuning were optimized using Gray Wolf Optimization (GWO) and Particle Swarm Optimization (PSO). Feature importance was assessed using SHAP values. Results: The proposed methodology achieved an accuracy of 85%, with a precision of 0.85, a recall of 0.89, and an F1 measure of 0.87, demonstrating the effectiveness of the approach in detecting and classifying rheumatoid nodules in both binary and multi-class classification scenarios. Conclusions: This study presents a robust tool for the detection and classification of nodules, particularly rheumatoid nodules, in medical imaging, offering significant potential for improving diagnostic accuracy and aiding in the early identification of rheumatoid conditions.

Spatial-Spectrum Two-Branch Model Based on a Superpixel Graph Convolutional Network and 1DCNN for Geochemical Anomaly Identification

In retail industry ensuring on shelf availability is crucial for customer satisfaction. This project introduces an innovative solution for automated on shelf stock monitoring by Graph Convolution Network (GCN) fundamental algorithm as framework. This particular system will combine computer vision techniques to capture real time shelf images and employees GCN to process the visual data efficiently. Through graph based representation of store shelves and products, the GCN algorithm will analyze the items and their availability status. This project includes in data collection, image preprocessing, object detection. The application of GCN is used to construct a dynamic graph that capture items. This algorithm’s ability to model complex dependencies within the shelf inventory facilitates accurate stock availability. Alerts are generated when low stock or out of stock situations are detected. Keywords—super pixel graph ,Graph convolutional network ,SVM , shelf image and notification alert

Researchers have been paying more attention to hyperspectral image (HSI) classification based on semi-supervised superpixel-level graph convolutional networks (SGCNs) due to their aggregation ability of rich contextual information. Although these SGCNs achieve good classification performance, the influence of the positional distributions among labeled superpixels has been overlooked. The locations of labeled superpixels, such as located at class boundaries or centers, exert a substantial influence on the final performance. To address this issue, this article proposed a novel graph neural network (GCN) method with the guidance of positional distributions of labeled superpixels, abbreviated as LPDGCN. Specifically, we first propose to utilize the sparse, low-rank as well as feature smoothness restrictions to optimize the initial superpixel graph structure because the connectivity relationships of labeled superpixels located at class boundaries or centers are easily influenced by spectral variation. Second, in order to effectively determine the positional distributions of labeled superpixels and make full use of the position relationships, we propose to utilize the information conflict from the above topology connectivity to determine the positional distributions of labeled superpixels and develop the reweighted strategy to weaken the influence of labeled superpixels located at class boundaries and strengthen the influence of that located at class centers. Finally, we evaluate the LPDGCN method on four public HSI datasets, demonstrating its superiority over other advanced classification methods in terms of three metrics, i.e., overall accuracy (OA), average accuracy (AA), and kappa coefficient (KC).

<abstract xmlns="http://www.w3.org/1999/xhtml"> Numerous image processing techniques have been developed for the identification of various types of skin lesions. In real-world scenarios, the specific lesion type is often unknown in advance, leading to a multi-class prediction challenge. The available evidence underscores the importance of employing a comprehensive array of diverse features and subsequently identifying the most important ones as a crucial step in visual diagnostics. For this purpose, we addressed both binary and five-class classification tasks using a small dataset, with skin lesions prevalent in Lithuania. The model was trained using a rich set of 662 features, encompassing both conventional image features and graph-based ones, which were obtained from the superpixel graph generated using Delaunay triangulation. We explored the influence of feature importance determined by SHAP values, resulting in a weighted F1-score of 92.48% for the two-class classification and 71.21% for the five-class prediction. </abstract>

Detecting infrared small targets (IRSTs) submerged in the complex back- grounds with heavy clutters is a challenging work of great significance in many fields. Inspired by representation capability of the superpixel, this paper presents a human visual system (HVS) based method called Homogeneity weighted local patch saliency (HWLPS). Firstly, the image is transformed into a superpixel graph and a compound flux field of the thermal intensity and gradient. Next, IRST component saliency and superpixel homogeneity are quantified by the morphing gradient direction diversity (GDD) metric and the inter-patch homogeneity (IPH) successively. Finally, the local contrast aggregation fuses the morphing GDD and the foreground saliency of image patches according to the superpixel graph. Experiments on various real IR image datasets show that, the proposed algorithm outperforms the state-of-the-art methods with robustness and stability against intricate back- grounds.

Prior studies using graph neural networks (GNNs) for image classification have focused on graphs generated from a regular grid of pixels or similar-sized superpixels. In the latter, a single target number of superpixels is defined for an entire dataset irrespective of differences across images and their intrinsic multiscale structure. On the contrary, this study investigates image classification using graphs generated from an image-specific number of multiscale superpixels. We propose WaveMesh, a new wavelet-based superpixeling algorithm, where the number and sizes of superpixels in an image are systematically computed based on its content. WaveMesh superpixel graphs are structurally different from similar-sized superpixel graphs. We use SplineCNN, a state-of-the-art network for image graph classification, to compare WaveMesh and similar-sized superpixels. Using SplineCNN, we perform extensive experiments on three benchmark datasets under three local-pooling settings: 1) no pooling, 2) GraclusPool, and 3) WavePool, a novel spatially heterogeneous pooling scheme tailored to WaveMesh superpixels. Our experiments demonstrate that SplineCNN learns from multiscale WaveMesh superpixels on-par with similar-sized superpixels. In all WaveMesh experiments, GraclusPool performs poorer than no pooling / WavePool, showing that poor cluster assignment negatively affects the performance of the network while learning from multiscale superpixels.

At present, hyperspectral image classification (HSIC) technology based on deep learning has been widely explored. However, the time and labor cost of obtaining enough labeled samples are expensive. To obtain higher classification performance with a few number of labeled samples, a coarse-to-fine semi-supervised classification learning (CFSSL) method is proposed in this letter. First of all, the CFSSL performs coarse-grained classification with a few number of labeled samples, and the breaking-ties (BT) criterion is introduced to sample the coarse-grained classification results to ensure that the samples with high confidence are selected to generate pseudo-labels. Then, the pseudo-labels and their corresponding unlabeled samples are sent to the feature extraction network for fine-grained classification, so as to obtain more advanced classification results. Finally, in the fine-grained classification stage, a multi-scale convolution kernel attention aggregation network ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">A</i> <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> -MCKN) is designed to simultaneously extract the spatial-spectral features of the image and ensure clear texture boundaries of ground objects. Experimental results on two public datasets show that the CFSSL can obtain better accuracy than other methods with a few number of labeled samples.

• We develop a water detection method which is particularly efficient in small water. • Graph convolutional network is innovatively used in image semantic segmentation. • Robustness of SG-waterNet is proven over urban, agricultural and mountainous scenes. Small waterbodies sustain susceptible ecosystems and are influenced by variable dynamics associated with human activities and environmental disturbances. Although remote sensing has displayed efficiency in mapping surface waterbodies on a regular basis, the identification of small waterbodies such as ponds or irrigation ditches remains a challenge, as small waterbodies are often confused with other low-reflectivity surfaces. In this study, a superpixel-based graph convolutional network (GCN) for small waterbody extraction (SG-waterNet) is proposed. Specifically, the SG-waterNet method includes a new object-based representation of an image called a superpixel graph. The superpixel graph contains compact spectral and contextual information and can be exploited by the GCN. A deep GCN architecture is used to efficiently preserve small waterbody features and detect surface waterbodies with high completeness and correctness. We tested the proposed approach on a frequently used open-access Gaofen Image Dataset (GID) and Gaofen-1 image from Hubei Province in China (a total of 11,660 km 2 for research). The extraction accuracy of SG-waterNet for small waterbodies (<2 ha) was between 84.31% and 89.77% at the five evaluation sites, and the method extracted waterbodies 300 m 2 and larger with high confidence. Compared with six state-of-the-art methods, SG-waterNet exhibited significant sensitivity to small waterbodies (especially smaller than 100 m 2 ) and detected small waterbody boundaries with the highest completeness and correctness. The average accuracy improvement achieved with SG-waterNet at the evaluation sites ranged from 11.10% to 13.87%. The proposed method is a significant advancement in small waterbody monitoring and can provide promising and practical solutions for real-world applications.

Hyperspectral image classification (HSIC) methods usually require more training samples for better classification performance. However, a large number of labeled samples are difficult to obtain because it is cost- and time-consuming to label an HSI in a pixel-wise way. Therefore, how to overcome the problem of insufficient accuracy and stability under the condition of small labeled training sample size (SLTSS) is still a challenge for HSIC. In this paper, we proposed a novel multiple superpixel graphs learning method based on adaptive multiscale segmentation (MSGLAMS) for HSI classification to address this problem. First, the multiscale-superpixel-based framework can reduce the adverse effect of improper selection of a superpixel segmentation scale on the classification accuracy while saving the cost to manually seek a suitable segmentation scale. To make full use of the superpixel-level spatial information of different segmentation scales, a novel two-steps multiscale selection strategy is designed to adaptively select a group of complementary scales (multiscale). To fix the bias and instability of a single model, multiple superpixel-based graphical models obatined by constructing superpixel contracted graph of fusion scales are developed to jointly predict the final results via a pixel-level fusion strategy. Experimental results show that the proposed MSGLAMS has better performance when compared with other state-of-the-art algorithms. Specifically, its overall accuracy achieves 94.312%, 99.217%, 98.373% and 92.693% on Indian Pines, Salinas and University of Pavia, and the more challenging dataset Houston2013, respectively.

Oil and gas reservoirs are well studied in petroleum engineering, using seismic data to estimate fluid flow and well placement. However, seismic data cannot capture fractures due to their scale, and fractures may affect rock porosity and permeability. Consequently, rock fracture segmentation and quantification from aerial images of analogous outcrops can input essential information into those studies. This paper presents a new method, named <i>interactive Forest Growing</i> (iFG), for fracture segmentation. The image is initially segmented into superpixels, defining a superpixel graph. The user selects seed superpixels, a path-cost threshold, and fractures are delineated by growing one optimum-path tree from each seed with path costs limited to the selected threshold. iFG considerably increases efficiency while reducing human effort in fracture segmentation compared to pixel-by-pixel manual annotation. We evaluate iFG with three specialists and against an <i>interactive Region Growing</i> (iRG) method using 15 images to measure bias in user interpretations, verify efficiency gain over a similar approach, and generate a dataset with consolidated annotation for future work. The experiments show that iFG reduces user interventions from 19% to 33% compared to iRG, users with more experience in fracture analysis complete segmentation 4-5 times faster, and segmentation effectiveness is independent of user experience since the average F1 scores between users uing both methods ranged from 0.966 to 0.979, allowing us to create a consolidated segmentation.

Hyperspectral image (HSI) classification is one of the major problems in the field of remote sensing. Particularly, graph-based HSI classification is a promising topic and has received increasing attention in recent years. However, graphs with pixels as nodes generate large size graphs, thus increasing the computational burden. Moreover, satisfactory classification results are often not obtained without considering spatial information in constructing graph. To address these issues, this study proposes an efficient and effective semi-supervised spectral-spatial HSI classification method based on sparse superpixel graph (SSG). In the constructed sparse superpixels graph, each vertex represents a superpixel instead of a pixel, which greatly reduces the size of graph. Meanwhile, both spectral information and spatial structure are considered by using superpixel, local spatial connection and global spectral connection. To verify the effectiveness of the proposed method, three real hyperspectral images, Indian Pines, Pavia University and Salinas, are chosen to test the performance of our proposal. Experimental results show that the proposed method has good classification completion on the three benchmarks. Compared with several competitive superpixel-based HSI classification approaches, the method has the advantages of high classification accuracy (&gt;97.85%) and rapid implementation (&lt;10 s). This clearly favors the application of the proposed method in practice.