Search Results (9)

Instance segmentation of 3D point clouds is a fundamental task for scene understanding in applications such as autonomous driving, robotics, and augmented reality. The inherent irregularity and sparsity of point clouds, compounded by scale variations and instance adhesion, pose significant challenges to accurate segmentation. Existing grouping-based methods are often limited by the loss of geometric details in single-path backbones and by error propagation near complex boundaries. To address these issues, a Multi-grained Dual-aware Grouping algorithm (MDGroup) is proposed, which explicitly integrates multi-grained feature representation with dual awareness of class and boundary. The algorithm features a Dual-Resolution 3D U-Net (DRNet) that preserves local geometric details while aggregating global semantics through adaptive alignment. A four-branch prediction scheme enhances semantic and offset estimation with boundary and directional cues, enabling fine-grained boundary modeling. Furthermore, a Hierarchical Adaptive Multi-grained Feature fusion framework (HAMF) achieves efficient cross-scale alignment by combining Class-Aware Dynamic Voxelization and Class-Aware Pyramid Scaling. Finally, a Boundary-Aware Weighted Aggregation mechanism (BAWA) refines instance grouping by dynamically weighting semantic confidence, geometric distance, boundary probability, and directional consistency. To extend the model to dynamic scenes, a Temporal Adaptive Gating (TAG) module is introduced to leverage historical frame correlations. Extensive experiments on the ScanNet v2, S3DIS, STPLS3D, SemanticKITTI, LiDAR-Net, and OCID benchmarks demonstrate that MDGroup achieves state-of-the-art performance among grouping-based methods, particularly on small objects, complex boundaries, and dynamic environments. Full article

Hyperedge prediction is crucial for uncovering higher-order relationships in complex systems but faces core challenges, including unmodeled node influence heterogeneity, overlooked hyperedge order effects, and data sparsity. This paper proposes Order propagation Fusion Self-supervised learning for Hyperedge prediction (OFSH) to address these issues. OFSH introduces a hyperedge order propagation mechanism that dynamically learns node importance weights and groups neighbor hyperedges by order, applying max–min pooling to amplify feature distinctions. To mitigate data sparsity, OFSH incorporates a key node-guided augmentation strategy with adaptive masking, preserving core high-order semantics. It identifies topological hub nodes based on their comprehensive influence and employs adaptive masking probabilities to generate augmented views preserving core high-order semantics. Finally, a triadic contrastive loss is employed to maximize cross-view consistency and capture invariant semantic information under perturbations. Extensive experiments on five public real-world hypergraph datasets demonstrate significant improvements over state-of-the-art methods in AUROC and AP. Full article

The effectiveness indicator system of remote sensing satellites includes various satellites capabilities. Effectiveness evaluation is the process of calculating these indicators in the digital world, involving many different physical parameters of multiple subsystems. Model-based simulation statistics method is the mainstream approach of effectiveness evaluation, and digital twin is currently the most advanced modeling method for simulation. The satellite digital twin model has the characteristics of multi-dynamic, multi-spatial scale and multi-physics field coupling, which gives rise to challenges related to the stiff problem of ordinary differential equations and multi-scale problem of partial differential equations to the calculation process of indicators. It is difficult to solve these problems by breakthroughs in numerical solution methods. This paper uses the sparsity of the satellite system to group each indicator of the effectiveness evaluation indicator system according to the change period. The satellite system model is decomposed into multiple modules according to the composition and structure, and a series of models with different simulation fidelity are established for each module. The optimization schemes for selecting model granularity when calculating indicators by group is given. Simulation results show that this approach considers the coupling between systems, grasps the main contradiction of indicator calculation and overcomes the loss of indicator accuracy caused by the separate calculation of each subsystem under the neglect of coupling in the traditional method. Additionally, it avoids the difficulty in numerical calculation caused by coupling, while simultaneously balancing the accuracy and efficiency of the model simulations. Full article

Pedestrians are often obstructed by other objects or people in real-world vision sensors. These obstacles make pedestrian-attribute recognition (PAR) difficult; hence, occlusion processing for visual sensing is a key issue in PAR. To address this problem, we first formulate the identification of non-occluded frames as temporal attention based on the sparsity of a crowded video. In other words, a model for PAR is guided to prevent paying attention to the occluded frame. However, we deduced that this approach cannot include a correlation between attributes when occlusion occurs. For example, “boots” and “shoe color” cannot be recognized simultaneously when the foot is invisible. To address the uncorrelated attention issue, we propose a novel temporal-attention module based on group sparsity. Group sparsity is applied across attention weights in correlated attributes. Accordingly, physically-adjacent pedestrian attributes are grouped, and the attention weights of a group are forced to focus on the same frames. Experimental results indicate that the proposed method achieved 1.18% and 6.21% higher $F_1$-scores than the advanced baseline method on the occlusion samples in DukeMTMC-VideoReID and MARS video-based PAR datasets, respectively. Full article

The construction of diverse dictionaries for sparse representation of hyperspectral image (HSI) classification has been a hot topic over the past few years. However, compared with convolutional neural network (CNN) models, dictionary-based models cannot extract deeper spectral information, which will reduce their performance for HSI classification. Moreover, dictionary-based methods have low discriminative capability, which leads to less accurate classification. To solve the above problems, we propose a deep learning-based structure dictionary for HSI classification in this paper. The core ideas are threefold, as follows: (1) To extract the abundant spectral information, we incorporate deep residual neural networks in dictionary learning and represent input signals in the deep feature domain. (2) To enhance the discriminative ability of the proposed model, we optimize the structure of the dictionary and design sharing constraint in terms of sub-dictionaries. Thus, the general and specific feature of HSI samples can be learned separately. (3) To further enhance classification performance, we design two kinds of loss functions, including coding loss and discriminating loss. The coding loss is used to realize the group sparsity of code coefficients, in which within-class spectral samples can be represented intensively and effectively. The Fisher discriminating loss is used to enforce the sparse representation coefficients with large between-class scatter. Extensive tests performed on hyperspectral dataset with bright prospects prove the developed method to be effective and outperform other existing methods. Full article

Recently, non-negative tensor factorization (NTF) as a very powerful tool has attracted the attention of researchers. It is used in the unmixing of hyperspectral images (HSI) due to its excellent expression ability without any information loss when describing data. However, most of the existing unmixing methods based on NTF fail to fully explore the unique properties of data, for example, low rank, that exists in both the spectral and spatial domains. To explore this low-rank structure, in this paper we learn the different low-rank representations of HSI in the spectral, spatial and non-local similarity modes. Firstly, HSI is divided into many patches, and these patches are clustered multiple groups according to the similarity. Each similarity group can constitute a 4-D tensor, including two spatial modes, a spectral mode and a non-local similarity mode, which has strong low-rank properties. Secondly, a low-rank regularization with logarithmic function is designed and embedded in the NTF framework, which simulates the spatial, spectral and non-local similarity modes of these 4-D tensors. In addition, the sparsity of the abundance tensor is also integrated into the unmixing framework to improve the unmixing performance through the $L_{2,1}$ norm. Experiments on three real data sets illustrate the stability and effectiveness of our algorithm compared with five state-of-the-art methods. Full article

Distance weighted discrimination (DWD) is an appealing classification method that is capable of overcoming data piling problems in high-dimensional settings. Especially when various sparsity structures are assumed in these settings, variable selection in multicategory classification poses great challenges. In this paper, we propose a multicategory generalized DWD (MgDWD) method that maintains intrinsic variable group structures during selection using a sparse group lasso penalty. Theoretically, we derive minimizer uniqueness for the penalized MgDWD loss function and consistency properties for the proposed classifier. We further develop an efficient algorithm based on the proximal operator to solve the optimization problem. The performance of MgDWD is evaluated using finite sample simulations and miRNA data from an HIV study. Full article

As artificial intelligence (AI)- or deep-learning-based technologies become more popular, the main research interest in the field is not only on their accuracy, but also their efficiency, e.g., the ability to give immediate results on the users’ inputs. To achieve this, there have been many attempts to embed deep learning technology on intelligent sensors. However, there are still many obstacles in embedding a deep network in sensors with limited resources. Most importantly, there is an apparent trade-off between the complexity of a network and its processing time, and finding a structure with a better trade-off curve is vital for successful applications in intelligent sensors. In this paper, we propose two strategies for designing a compact deep network that maintains the required level of performance even after minimizing the computations. The first strategy is to automatically determine the number of parameters of a network by utilizing group sparsity and knowledge distillation (KD) in the training process. By doing so, KD can compensate for the possible losses in accuracy caused by enforcing sparsity. Nevertheless, a problem in applying the first strategy is the unclarity in determining the balance between the accuracy improvement due to KD and the parameter reduction by sparse regularization. To handle this balancing problem, we propose a second strategy: a feedback control mechanism based on the proportional control theory. The feedback control logic determines the amount of emphasis to be put on network sparsity during training and is controlled based on the comparative accuracy losses of the teacher and student models in the training. A surprising fact here is that this control scheme not only determines an appropriate trade-off point, but also improves the trade-off curve itself. The results of experiments on CIFAR-10, CIFAR-100, and ImageNet32 × 32 datasets show that the proposed method is effective in building a compact network while preventing performance degradation due to sparsity regularization much better than other baselines. Full article

Image recovery from compressive sensing (CS) measurement data, especially noisy data has always been challenging due to its implicit ill-posed nature, thus, to seek a domain where a signal can exhibit a high degree of sparsity and to design an effective algorithm have drawn increasingly more attention. Among various sparsity-based models, structured or group sparsity often leads to more powerful signal reconstruction techniques. In this paper, we propose a novel entropy-based algorithm for CS recovery to enhance image sparsity through learning the group sparsity of residual. To reduce the residual of similar packed patches, the group sparsity of residual is described by a Laplacian scale mixture (LSM) model, therefore, each singular value of the residual of similar packed patches is modeled as a Laplacian distribution with a variable scale parameter, to exploit the benefits of high-order dependency among sparse coefficients. Due to the latent variables, the maximum a posteriori (MAP) estimation of the sparse coefficients cannot be obtained, thus, we design a loss function for expectation–maximization (EM) method based on relative entropy. In the frame of EM iteration, the sparse coefficients can be estimated with the denoising-based approximate message passing (D-AMP) algorithm. Experimental results have shown that the proposed algorithm can significantly outperform existing CS techniques for image recovery. Full article