Robust object detection based on local similar structure statistical matching

Molecular simulation is central to modern drug discovery but is often limited by high computational cost and the complexity of molecular interactions. Deep-learning drug–target interaction (DTI) prediction can accelerate screening; however, many models underuse the local functional structure features—binding motifs, reactive groups, and residue-level fragments—that drive recognition. We present LoF-DTI, a framework that explicitly represents and couples such local features. Drugs are converted from SMILES into molecular graphs and targets from sequences into feature representations. On the drug side, a Jumping Knowledge (JK) enhanced Graph Isomorphism Network (GIN) extracts atom- and neighborhood-level patterns; on the target side, residual CNN blocks with progressively enlarged receptive fields, augmented by N-mer substructural statistics, capture multi-scale local motifs. A Gated Cross-Attention (GCA) module then performs atom-to-residue interaction learning, highlighting decisive local pairs and providing token-level interpretability through attention scores. By prioritizing locality during both encoding and interaction, LoF-DTI delivers competitive results across multiple benchmarks and improves early retrieval relevant to virtual screening. Case analyses show that the model recovers known functional binding sites, suggesting strong potential to provide mechanism-aware guidance for molecular simulation and to streamline the drug design pipeline.

Accurate and robust object detection is of critical importance in airport surface surveillance to ensure the security of air transportation systems. Owing to the constraints imposed by a relatively fixed receptive field, existing airport surface detection methods have not yet achieved substantial advancements in accuracy. Furthermore, these methods are vulnerable to adversarial attacks with carefully crafted adversarial inputs. To address these challenges, we propose the Vision GNN-Edge (ViGE) block, an enhanced block derived from the Vision GNN (ViG). ViGE introduces the receptive field in pixel space and represents the spatial relation between pixels directly. Moreover, we implement an adversarial training strategy with augmented training samples generated by adversarial perturbation. Empirical evaluations on the public remote sensing dataset LEVIR and a manually collected airport surface dataset show that: 1. our proposed method surpasses the original model in precision and robustness; 2. defining the receptive field in pixel space performs better than that on representation space.

A method of real-time object detection using Radial Reach Correlation (RRC) has been developed. We also apply a statistical background estimation to cope with dynamic and complex environments. In order to reduce the computational cost of RRC for an embedded image processing device, we apply two techniques: fast RRC algorithm and background estimation based on statistical approach with cumulative averaging process. As a result, it is possible to reduce the RRC calculation process to about 1/4 in comparison with normal RRC. By combining these techniques, we can achieve robust real-time object detection.

Existing methods of cross-modality person re-identification usually use set-level global feature constraints to reduce the cross-modality discrepancy. However they often ignore the modality-specific and modality-shared local feature matching. Modality-specific local matching could help extract discriminative identity-consistent features and alleviate spatial misalignment, and modality-shared local matching could make identity-consistent features be correlated in two modalities and mitigate the modality misalignment. Therefore, we establish the intra-modality and inter-modality co-occurrence relation between identity parts. We observe the modality-specific and modality-shared local feature matching of identity parts as intra-modality and inter-modality low-rank relation-finding problems. Therefore, in this work, we propose low-rank local matching (LLM) approach to establish the intra-modality and inter-modality co-occurrence relation between identity parts. First, in order to reinforce identity-consist features in two modalities and correlate these cross-modality identity-consist features together, the local matching (LM) module is designed to estimate partial co-occurrence probability by measuring local feature similarity. Moreover, we propose cross-modality triplet-center loss (CTLoss) which adds global constraint for each class distribution in embedding space to alleviate dramatic data expansion due to modality variance. Extensive experiments on two datasets demonstrate the superior performance of our approach over the existing state-of-the-art. Our code is released on https://github.com/fegnyujian/LLM .

Using methods from algebraic graph theory and convex optimization, we study the relationship between local structural features of a network and the eigenvalues of its Laplacian matrix. In particular, we propose a series of semidefinite programs to find new bounds on the spectral radius and the spectral gap of the Laplacian matrix in terms of a collection of local structural features of the network. Our analysis shows that the Laplacian spectral radius is strongly constrained by local structural features. On the other hand, we illustrate how local structural features are usually insufficient to accurately estimate the Laplacian spectral gap. As a consequence, random graph models in which only local structural features are prescribed are, in general, inadequate to faithfully model Laplacian spectral properties of a network.

This paper deals with the retrieval of scene-of-cri me (or scene) shoeprint images from a reference database of shoeprint images by using a n ew local feature detector and an improved local feature descriptor. Our approach is based on novel modifications and improvements of a few recent techniques in this area: (1) the scale a dapted Harris detector, which is an extension to multi-scale domains of the Harris corner detector; (2) automatic scale selection by the characteristic scale of a local structure. (3) SIFT (Scale-Invariant Feature Transform), one of the most widely investigated descriptors in recent year s. Like most of other local feature representations, the proposed approach can also be divided into two stages: (i) a set of distinctive local features are selected by first de tecting scale adaptive Harris corners where each of them is associated with a scale factor, and then selecting as the final result only those corners whose scale matches the scale of blob-like structur es around them. Here, the scale of a blob-like structure is detected by the Laplace-based scale se lection. (ii). for each feature detected, an enhanced SIFT descriptor is computed to represent t his feature. Our improvements lead two novel methods which we call the Modified Harris-Laplace (MHL) detector, and the enhanced SIFT descriptor. In this paper, we demonstrate the application of t he proposed scheme to the shoeprint image retrieval problem using six sets of synthetic scene images, 50 images for each, and a database of 500 reference shoeprint images. The ret rieval performance of the proposed approach is significantly better, in terms of cumul ative matching score, than the existing methods used in this application area, such as edge directional histogram, power spectral distribution, and pattern & topological spectra.

The diesel engine fault data is nonlinear and it’s difficult to extract the characteristic information. Kernel Principal Component Analysis (KPCA) is used to extract features of nonlinear data, only considering global structure. Kernel Locality Preserving Projection (KLPP) considers the local feature structure. So an improved algorithm for global and local structure preserving is proposed to extract the feature of data. The improved feature extraction algorithm combining KPCA and KLPP, avoids the loss of information considering the global and local feature structure and then uses the modified K-NN algorithm for fault classification. In this paper, the software AVL BOOST is used to simulate the faults of diesel engine. The simulation experiments indicate the proposed method can extract the feature vectors effectively, and shows good classification performance.

A CNN model embedded with local feature knowledge and its application to time-varying signal classification

Contemporary literature suggests that preattentive texture or pattern discrimination is induced by differences between local structure features or "textons." This paper presents a model for the description of such local structure features based on the computation of local autocorrelations within the image. By means of this structure model a measure of structure dissimilarity is introduced. Experiments have been carried out to test a hypothesized relation between the detectability of a target pattern in a field of background patterns and the value of the structure dissimilarity measure. The experimental results show that it seems justified to relate, in a quantitative way, the detectability of the target pattern to the value of the structure dissimilarity measure.

Decision letter: Visual and motor signatures of locomotion dynamically shape a population code for feature detection in Drosophila

Using deformable models to register medical images can result in problems of initialization of deformable models and robustness and accuracy of matching of inter-subject anatomical variability. To tackle these problems, a novel model is proposed in this paper by compounding local invariant features and global deformable geometry. This model has four steps. First, a set of highly-repeatable and highly-robust local invariant features, called Key Features Model (KFM), are extracted by an effective matching strategy. Second, local features can be matched more accurately through the KFM for the purpose of initializing a global deformable model. Third, the positional relationship between the KFM and the global deformable model can be used to precisely pinpoint all landmarks after initialization. And fourth, the final pose of the global deformable model is determined by an iterative process with a lower time cost. Through the practical experiments, the paper finds three important conclusions. First, it proves that the KFM can detect the matching feature points well. Second, the precision of landmark locations adjusted by the modeled relationship between KFM and global deformable model is greatly improved. Third, regarding the fitting accuracy and efficiency, by observation from the practical experiments, it is found that the proposed method can improve % of the fitting accuracy and reduce around 50% of the computational time compared with state-of-the-art methods.

By an increase in the volume of image data in the age of information, a need for image categorization systems is greatly felt. Recent activities in this area has shown that the image description by local features, often has very strong similarities among local partials of an image, but these methods are also challenging, because the use of a set of vectors for each image does not permit the direct use of most of the common learning methods and distance functions. On the other hand, measuring the created similarities of the collection of unordered features is also problematic, because most of the proposed methods have rather high time complexities computationally. In this article, an unsupervised learning method of categorization of objects from a collection of unlabeled images is introduced. Each image is described by a set of unordered local features and clustering is performed on the basis of partial similarities existing among these sets of features. For this purpose, by the use of pyramid match algorithm, the set of the features are mapped in multi-resolution histograms and two sets of feature vectors in time-line are calculated in this new distance. These similarities are employed as a criterion distance among the patterns in hierarchical clustering; therefore, the categorization of objects by the use of common learning methods is performed with acceptable accuracy and faster than the existing algorithms.

Three-dimensional object recognition is complicated by the fact that an object may present itself to a sensor in any one of an infinity of poses. Fortunately, this problem can be alleviated by declaring equivalent classes of object poses through either the use of the aspect graph representation of objects [7, 14] or through the use of local feature sets (LFSs) [3, 9]. To elaborate a bit on the notion of local feature sets, since they are so central to the work described in this chapter, consider the two simple model objects shown in Figure 2.l(a) and 2.l(b). Each surface of each of these objects is called a feature and each feature can be characterized by a set of attribute-value pairs. For example, the surface features 9 and 4 of an instance of the object in Figure 2.l(a) can be described by the attribute-value pairs: How attributes are used in matching a feature from the scene with a feature from a model depends on the nature of the attribute. If an attribute, such as “area,” is susceptible to occlusion, then it can be used only in the sense of its providing an upper-bound constraint, meaning that we must exclude all model features whose areas exceed the area of the feature in the scene. This point is discussed in greater detail in [8].

Operating performance assessment of industrial process based on MIC-graph convolutional networks with local slow feature analysis.

In content based image retrieval (CBIR) process, every image has been represented in a compact set of local visual features i.e. color, texture, and/or shape of images. This set of local visual features is known as feature vector. In the CBIR process, feature vectors of images have been used to represent or to identify similar images in adequate way. As a result, feature vector construction has always been considered as an important issue since it must reflect proper image semantics using minimal amount of data. The proposed CBIR scheme is based on the combination of color and texture features. In this work initially, we have converted the given RGB image into HSV color image. Subsequently, we have considered H (hue), S (saturation), and V (intensity) components for extraction of visual image features. The texture features have been extracted from the V component of the image using 2D dual-tree complex wavelet transform (2D DT-CWT) where it analyzes the textural patterns in six different directions i.e. ±15\(^{\circ }\), ±45\(^{\circ }\), and ±75\(^{\circ }\). At the same time, we have computed the probability histograms of H and S components of the image respectively and subsequently those are divided into non-uniform bins based on cumulative probability for extraction of color based features. So, in this work both the color and texture features have been extracted simultaneously. Finally, the obtained features have been concatenated to attain the final feature vector and same is considered in image retrieval process. We have tested the novelty and performance of the proposed work in two Corel, two objects, and, a texture image datasets. The experimental results reveal the acceptable retrieval performances for different types of datasets.