Object Features Research Articles

Inter-object discriminative graph modeling for indoor scene recognition

Variable scene layouts and coexisting objects across scenes make indoor scene recognition still a challenging task. Leveraging object information within scenes to enhance the distinguishability of feature representations has emerged as a key approach in this domain. Currently, most object-assisted methods use a separate branch to process object information, combining object and scene features heuristically. However, few of them pay attention to interpretably handling the hidden discriminative knowledge within object information. In this paper, we propose to leverage discriminative object knowledge to enhance scene feature representations. Initially, we capture the object-scene discriminative relationships from a probabilistic perspective, which are transformed into an Inter-Object Discriminative Prototype (IODP). Given the abundant prior knowledge from IODP, we subsequently construct a Discriminative Graph Network (DGN), in which pixel-level scene features are defined as nodes and the discriminative relationships between node features are encoded as edges. DGN aims to incorporate inter-object discriminative knowledge into the image representation through graph convolution and mapping operations (GCN). With the proposed IODP and DGN, we obtain state-of-the-art results on several widely used scene datasets, demonstrating the effectiveness of the proposed approach.

Detection of Novel Objects without Fine-Tuning in Assembly Scenarios by Class-Agnostic Object Detection and Object Re-Identification

Object detection is a crucial capability of autonomous agents for human–robot collaboration, as it facilitates the identification of the current processing state. In industrial scenarios, it is uncommon to have comprehensive knowledge of all the objects involved in a given task. Furthermore, training during deployment is not a viable option. Consequently, there is a need for a detector that is able to adapt to novel objects during deployment without the necessity of retraining or fine-tuning on novel data. To achieve this, we propose to exploit the ability of discriminative embeddings learned by an object re-identification model to generalize to unknown categories described by a few shots. To do so, we extract object crops with a class-agnostic detector and then compare the object features with the prototypes of the novel objects. Moreover, we demonstrate that the embedding is also effective for predicting regions of interest, which narrows the search space of the class-agnostic detector and, consequently, increases processing speed. The effectiveness of our approach is evaluated in an assembly scenario, wherein the majority of objects belong to categories distinct from those present in the training datasets. Our experiments demonstrate that, in this scenario, our approach outperforms the current best few-shot object-detection approach DE-ViT, which also does not perform fine-tuning on novel data, in terms of both detection capability and inference speed.

Enhancing image caption generation through context-aware attention mechanism

Image captioning, the process of generating natural language descriptions based on image content, has garnered attention in AI research for its implications in scene understanding and human-computer interaction. While much prior research has focused on caption generation for English, addressing low-resource languages like Bengali presents challenges, particularly in producing coherent captions linking visual objects with corresponding words. This paper proposes a context-aware attention mechanism over semantic attention to accurately diagnose objects for image captioning in Bengali. The proposed architecture consists of an encoder and a decoder block. We chose ResNet-50 over the other pre-trained models for encoding the image features due to its ability to solve the vanishing gradient problem and recognize complex object features. For decoding generated captions, a bidirectional Gated Recurrent Unit (GRU) architecture combined with an attention mechanism captures contextual dependencies in both directions, resulting in more accurate captions. The paper also highlights the challenge of transferring knowledge between domains, especially with culturally specific images. Evaluation of three Bengali benchmark datasets, namely BAN-Cap, BanglaLekhaImageCaption, and Bornon, demonstrates significant performance improvement in METEOR score over existing methods by approximately 30%, 18%, and 45%, respectively. The proposed context-aware, attention-based image captioning system significantly outperforms current state-of-the-art models in Bengali caption generation despite limitations in reference captions on certain datasets.

Developing a methodology for user‐oriented verification of polar low forecasts

AbstractPolar lows exhibit features with very sharp weather contrasts. In weather forecasting, a small misplacement of areas with hazardously high wind speeds can have fatal impacts for people living in polar regions. Therefore, a novel application of spatial verification methods for objective metrics of size, shape, and location of areas with hazardous weather is tested. To separate the effect of errors in polar low location and direction of motion from errors relative to the polar low centre, surface wind fields from the limited‐area weather forecasting model Applications of Research to Operations at Mesoscale‐Arctic and Copernicus Climate Change Service Arctic Regional Reanalysis are centred at the polar low centre and rotated according to the direction of background flow surrounding the polar low. Then the possibilities of the features‐based verification methods SAL (structure, amplitude, location) and MODE (Method for Object‐based Diagnostic Evaluation) are explored using a test case from October 2019. The study demonstrates that the methodology can provide valuable information about forecast performance. MODE is a flexible method with metrics that focus on characteristics of individual objects and can be adapted to questions at hand. For example, a measure of storm eye size was added. SAL, on the other hand, provides effective summary metrics for the full domain and proved particularly useful for evaluation of the overall distribution of wind speed. To evaluate the number of correctly or incorrectly identified areas with harsh weather rather than their details about their shape, contingency scores are more suitable. Applied to a larger dataset, this methodology can assess performance as a function of forecast length, as well as geographical area, and the type of polar low. The methodology can also be applied to other types of low‐pressure systems, such as extratropical cyclones.

UAV Visual Tracking with Enhanced Feature Information

Unmanned aerial vehicles (UAVs) visual tracking is an important research direction. The tracking object is lost due to the problems of target occlusion, illumination variation, flight vibration and so on. Therefore, based on a Siamese network, this study proposes a UAVs visual tracker named SiamDFT++ to enhance the correlation of depth features. First, the network width of the three-layer convolution after the full convolution neural network is doubled, and the appearance information of the target is fully utilized to complete the feature extraction of the template frame and the detection frame. Then, the attention information fusion module and feature deep convolution module are proposed in the template branch and the detection branch, respectively. The feature correlation calculation methods of the two depths can effectively suppress the background information, enhance the correlation between pixel pairs, and efficiently complete the tasks of classification and regression. Furthermore, this study makes full use of shallow features to enhance the extraction of object features. Finally, this study uses the methods of deep cross-correlation operation and complete intersection over union to complete the matching and location tasks. The experimental results show that the tracker has strong robustness in UAVs short-term tracking scenes and long-term tracking scenes.

The laboratory as a pedagogical resource for teaching mathematics: the state of knowledge in ENEM communications

The development of mathematical learning begins through the relationship between the subject and the mathematical object with the use of concrete materials. In view of this, it is believed that the Mathematics Teaching Laboratory (LEM) is an environment that can promote mathematical learning through this relationship. Therefore, this qualitative research sought to understand how LEM can collaborate in the teaching and learning process of Basic Education students. To this end, a survey of articles published in editions of the Annals of ENEM’s (1987 – 2022) that had LEM as its theme was carried out; Subsequently, the texts were read in full, in which ten scientific communications were selected and constituted the object of analysis of this investigation. As main results, we highlight that LEM encourages students to develop activities using concrete and manipulable materials that allow them to review, learn about and deepen characteristics of objects of knowledge related to mathematics, developing interest in the study through the relationship that can be established with the social context to which this subject belongs.

MS23D: A 3D object detection method using multi-scale semantic feature points to construct 3D feature layer

LiDAR point clouds can effectively depict the motion and posture of objects in three-dimensional space. Many studies accomplish the 3D object detection by voxelizing point clouds. However, in autonomous driving scenarios, the sparsity and hollowness of point clouds create some difficulties for voxel-based methods. The sparsity of point clouds makes it challenging to describe the geometric features of objects. The hollowness of point clouds poses difficulties for the aggregation of 3D features. We propose a two-stage 3D object detection framework, called MS23D. (1) We propose a method using voxel feature points from multi-branch to construct the 3D feature layer. Using voxel feature points from different branches, we construct a relatively compact 3D feature layer with rich semantic features. Additionally, we propose a distance-weighted sampling method, reducing the loss of foreground points caused by downsampling and allowing the 3D feature layer to retain more foreground points. (2) In response to the hollowness of point clouds, we predict the offsets between deep-level feature points and the object’s centroid, making them as close as possible to the object’s centroid. This enables the aggregation of these feature points with abundant semantic features. For feature points from shallow-level, we retain them on the object’s surface to describe the geometric features of the object. To validate our approach, we evaluated its effectiveness on both the KITTI and ONCE datasets.

Geometric-aware RGB-D representation learning for hand–object reconstruction

Reconstructing hand–object interaction from single images, crucial for interactive applications, is challenging due to the diversity of hand–object poses and shapes. Depth maps effectively complement RGB data in understanding these interactions geometrically within challenging scenes. However, most existing methods do not fully utilize the potential benefits of RGB-D information fused at different feature levels for reconstruction, limiting their capability to capture the geometric details of hand–object interaction. To address this, we propose an implicit geometric-aware RGB-D representation learning approach using adaptive bidirectional RGB-D feature fusion (ABF) and geometric Fourier feature encoding (GFFE). This method innovatively leverages the hierarchical and complementary high-level features of RGB-D information to enhance the neural implicit representations of hand–object interaction during the feature extraction and encoding stages of RGB-D data. Initially, a two-stream RGB-point cloud encoder, combining CNN and Transformer architectures, extracts appearance and geometric information from RGB and point cloud data. Subsequently, ABF fuses this information, generating dense RGB-D features by leveraging the complementary of appearance and geometric features along with the varying sensitivities of different layers. Finally, GFFE introduces frequency domain information to capture both high-frequency and low-frequency features of objects equally to precisely capture the geometric details of hand–object interaction. Experiments conducted on the real-world DexYCB and the synthetic ObMan benchmarks demonstrate that our approach significantly outperforms existing approaches.

Digital technologies for signal radio vision and radio monitoring

Objectives. Radiophysical processes involving the electrodynamic formation of signal radio images diffusely scattered by the signature of small-sized objects or induced by the near field of radio devices are relevant for identifying radiogenomic (cumulant) features of objects in the microwave range in the development of neuroimaging ultra-short pulse (USP) signal radio vision systems, telemonitoring, and near-radio detection. The paper sets out to develop methods and algorithms for vector analysis of radio wave deformation of nonstationary fields forming a signal radio image based on radiophysical and topological characteristics of small-sized objects; to develop software and hardware for registration and neural network recognition of signal radio images, including methods for the synthesis and extraction of signal radiogenomes using digital twins of objects obtained through vector electrodynamic modeling; and to analyze signal radio images induced by elements of printed topology of electronic devices.Methods. The study is based on statistical radiophysics methods, time-frequency approaches for wavelet transformation of USP radio images, numerical electrodynamic methods for creating digital twins of small-sized objects, as well as neural network authentication algorithms based on the cumulant theory of pole-genetic and resonant physically unclonable functions used in identifying signal radio images.Results. The results of fundamental research on electrodynamic effects of vector-wave deformation of nonstationary fields of sub-nanosecond configuration are presented as a means of identifying and authenticating signal radio images. Neural network techniques for cumulant formation of radio genomes of signal radio images are proposed on the basis of pole-genetic and resonant functions.Conclusions. A radiogenome, representing the unique authenticator of a radio image, is shown to be formed on the basis of physically unclonable functions determined by the structure and set of radiophysical parameters of the image. Cumulant features of signal radio images identified on the basis of pole-genetic and physically unclonable resonant functions of small-sized objects are revealed.

Recognition of Organic Waste Objects Based on Vision Systems Using Attention Convolutional Neural Networks Models

Purpose: High population growth and increasing consumption patterns have resulted in significant organic waste production. The public often does not understand the correct way to deal with the problem of organic waste, including public awareness regarding the need for its management. Therefore, a system is needed to recognize waste objects based on various types. Currently, much research in this field has been studying object recognition, for example, the implementation of the Convolutional Neural Networks (CNN) model. However, there are still various challenges that must be addressed, including objects with diverse visual characteristics such as form, size, color, and physical condition. This research focuses on developing a system that enhances object recognition of waste, specifically organic waste, using an Attention Convolutional Neural Network (ACNN). By integrating attention mechanisms into the CNN model, this study addresses the challenges of recognizing waste objects with diverse visual characteristics. The proposed system seeks to improve the accuracy and efficiency of organic waste identification, which is crucial for advancing waste management practices and reducing environmental impact. Methods: This research combines a CNN architecture with an attention mechanism to create a better object detection environment called Attention-CNN (ACNN). The ACNN architecture employed consists of one layer input, three convoluted layers, three max-pooling layers, one attention layer, one flattened layer, four dropout layers, and two dense layers arranged in a certain way. Result: The research result shows that the model CNN with attention mechanism (ACNN) was slightly better at 86.93% than the standard model of CNN, which accounted for 86.70% in accuracy. Novelty: In general, the current use of CNN architecture to address waste object recognition problems typically employs standard architectures, resulting in lower accuracy for complex waste objects. In contrast, our research integrates attention mechanisms into the CNN architecture (ACNN), enhancing the model's ability to focus on relevant features of waste objects. This leads to improved recognition accuracy and robustness against visual variability. This distinction is important as it overcomes the limitations of standard CNN models in handling visually diverse and complex waste objects, thereby highlighting the novelty and contribution of our research.

HSC: a multi-hierarchy descriptor for loop closure detection in overhead occlusion scenes

Loop closure detection is a key technology for robotic navigation. Existing research primarily focuses on feature extraction from global scenes but often neglects local overhead occlusion scenes. In these local scenes, objects such as vehicles, trees, and buildings vary in height, creating a complex multi-layered structure with vertical occlusions. Current methods predominantly employ a single-level extraction strategy to construct descriptors, which fails to capture the characteristics of occluded objects. This limitation results in descriptors with restricted descriptive capabilities. This paper introduces a descriptor named Hierarchy Scan Context (HSC) to address this shortfall. HSC effectively extracts height feature information of objects at different levels in overhead occlusion scenes through hierarchical division, demonstrating enhanced descriptive capabilities. Additionally, a time series enhancement strategy is proposed to reduce the number of algorithmic missed detections. In the experiments, the proposed method is validated using a self-collected dataset and the public KITTI and NCLT datasets, demonstrating superior performance compared to competitive methods. Furthermore, the proposed method also achieves an average maximum F1 score of 0.92 in experiments conducted on nine selected road segments with overhead occlusion.

Two "What" Networks in the Human Brain.

Ungerleider and Mishkin, in their influential work that relied on detailed anatomical and ablation studies, suggested that visual information is processed along two distinct pathways: the dorsal "where" pathway, primarily responsible for spatial vision, and the ventral "what" pathway, dedicated to object vision. This strict division of labor has faced challenges in light of compelling evidence revealing robust shape and object selectivity within the putative "where" pathway. This article reviews evidence that supports the presence of shape selectivity in the dorsal pathway. A comparative examination of dorsal and ventral object representations in terms of invariance, task dependency, and representational content reveals similarities and differences between the two pathways. Both exhibit some level of tolerance to image transformations and are influenced by tasks, but responses in the dorsal pathway show weaker tolerance and stronger task modulations than those in the ventral pathway. Furthermore, an examination of their representational content highlights a divergence between the responses in the two pathways, suggesting that they are sensitive to distinct features of objects. Collectively, these findings suggest that two networks exist in the human brain for processing object shapes, one in the dorsal and another in the ventral visual cortex. These studies lay the foundation for future research aimed at revealing the precise roles the two "what" networks play in our ability to understand and interact with objects.

Возможности метода фокус-групп в процессе подготовки специалистов в сфере социальной работы

The main advantages and disadvantages of the focus group method as a qualitative method of em-pirical research are revealed. The main directions of application of the focus group method are de-fined - the effectiveness of general and private social work technologies, changes in the main char-acteristics of objects and subjects of social work, the development of new forms, technologies, di-rections of social work, the implementation of social work in the non–commercial sector. It is con-cluded that the focus group method is actively used in such areas of social work as education, healthcare, and social protection, and the ability to apply it in practice reflects a high level of train-ing of specialists and a deep understanding of the interdisciplinary connections between sociology and social work.

Multidomain Object Detection Framework Using Feature Domain Knowledge Distillation.

Object detection techniques have been widely studied, utilized in various works, and have exhibited robust performance on images with sufficient luminance. However, these approaches typically struggle to extract valuable features from low-luminance images, which often exhibit blurriness and dim appearence, leading to detection failures. To overcome this issue, we introduce an innovative unsupervised feature domain knowledge distillation (KD) framework. The proposed framework enhances the generalization capability of neural networks across both low- and high-luminance domains without incurring additional computational costs during testing. This improvement is made possible through the integration of generative adversarial networks and our proposed unsupervised KD process. Furthermore, we introduce a region-based multiscale discriminator designed to discern feature domain discrepancies at the object level rather than from the global context. This bolsters the joint learning process of object detection and feature domain distillation tasks. Both qualitative and quantitative assessments shown that the proposed method, empowered by the region-based multiscale discriminator and the unsupervised feature domain distillation process, can effectively extract beneficial features from low-luminance images, outperforming other state-of-the-art approaches in both low- and sufficient-luminance domains.

3D LoD2 modelling of Halong City based on UAV point cloud

3D urban building models are crucial for linking, converging, and integrating economic and social urban data. They are extensively utilized in numerous domains such as smart city development, comprehensive social management, and emergency decision-making. Advanced technologies like affordable UAV (Unmanned Aerial Systems) imagery enable a greater level of automation in data acquisition compared to traditional digital photogrammetry methods. The main objective of this research is to build a 3D LoD2 CityGML model with dense point clouds from UAV images. This study presents a completed workflow for generating 3D CityGML models of the city at LoD2 using UAV data. The use of dense point cloud data from UAV technology in the experimental area has been performed using a DJI Phantom 4 Pro. The original point clouds should be denoised using the statistical outlier remover (SOR), the main goal is to reduce noise while preserving the building's geometry. After that, the point clouds of object features were vectored for the level of detail 2 (LoD2) of the object's 3D volume corresponding to its actual height, and objects belonging to the Feature Class 3D layer will represent LoD2 on SketchUp Pro 2021 software to generate a highly accurate 3D model. The evaluation results show that the square errors calculated from the test points for the three axes X, Y, Z are 1.4 cm, 1.6 cm, 1.7 cm, respectively. Conducting research to choose the UAV photography method aims to offer an efficient and cost-effective solution, saving time and human resources, to address the 3D mapping challenges in urban areas across Vietnam.

A novel underwater object detection enhanced algorithm based on YOLOv5‐MH

AbstractUnderwater object detection is an important application of underwater vehicles. However, traditional underwater object detection algorithms have several shortcomings in underwater settings. These include imprecise feature extraction, slow detection speeds, and a lack of robustness. To address these shortcomings, a novel approach based on YOLOv5‐MH (Multi‐Head) is proposed in this paper. Firstly, an image enhancement technique is utilised. This technique uses adaptive linear mapping to adjust contrast and improve the quality of underwater images. Secondly, the C2f module for feature extraction is employed to enable more effective capture of object characteristics. Subsequently, the multi‐head self‐attention and coordinate attention are integrated into the network's backbone. This integration increases the attention given to input data, thereby enhancing the network's performance in handling complex tasks. Furthermore, a bidirectional feature pyramid is implemented to adeptly handle objects of varying scales and sizes, and elevate model performance. Finally, through comprehensive testing on the 2018 URPC dataset and deep plastic dataset, this method demonstrates superior performance. This performance is compared to the original YOLOv5 and other similar networks. It holds immense promise for practical applications across a wide spectrum of underwater tasks.

A Diffusion Neural Network-Enhanced Object Tracking Approach Under Sports Scenarios

Object tracking for motion scenes is a common research concern in field of computer vision. Its goal is to accurately track targets in different time periods and predict their future states by utilizing the motion information in video sequences. However, traditional target-tracking methods in motion scenes often face challenges such as target blur, occlusion, and changes in lighting. To deal with this issue, this paper proposes a diffusion neural network-enhanced object-tracking approach under sports scenarios. In order to further improve tracking performance, the diffusion convolution operation is introduced, which propagates features at different time steps to enhance the modeling ability of target motion. Then, suitable influencing factors are selected based on motion scene object feature parameters. Finally, a target tracking method is established by integrating these two methods. In the experiment, we used a large number of real motion scene datasets to evaluate the proposed method. The experimental results show that by comparing with traditional moving object tracking methods, the proposal achieves significant improvement in tracking accuracy and methodology robustness. In addition, we also conducted stability experiments, proving that this method has good stability for models with varying kernel numbers.

A monocular visual body enhancement algorithm for recreating simulation training games for sports students on the field

At present, the simulation training game for sports students has become an important auxiliary tool in the field of sports training. However, how to provide a more authentic and immersive experience has always been a hot and difficult research topic. In view of this, this paper proposes a method of visual sense enhancement based on monocular vision. In order to improve the accuracy of motion information extraction, Scale-invariant feature transform and sliding window filtering algorithm were used to optimize the results of motion information extraction. The experimental results show that the displacement acceleration and angular velocity on each axis vary significantly from frame to frame, for example, the Y-axis angular velocity varies from 0.178 to 0.924. The monocular vision enhancement algorithm can effectively reflect the dynamic characteristics of objects in the process of motion, help to simulate and reproduce the real motion of things on the field more accurately, and help to improve the skill level of athletes in the training process.

HP-YOLOv8: High-Precision Small Object Detection Algorithm for Remote Sensing Images.

YOLOv8, as an efficient object detection method, can swiftly and precisely identify objects within images. However, traditional algorithms encounter difficulties when detecting small objects in remote sensing images, such as missing information, background noise, and interactions among multiple objects in complex scenes, which may affect performance. To tackle these challenges, we propose an enhanced algorithm optimized for detecting small objects in remote sensing images, named HP-YOLOv8. Firstly, we design the C2f-D-Mixer (C2f-DM) module as a replacement for the original C2f module. This module integrates both local and global information, significantly improving the ability to detect features of small objects. Secondly, we introduce a feature fusion technique based on attention mechanisms, named Bi-Level Routing Attention in Gated Feature Pyramid Network (BGFPN). This technique utilizes an efficient feature aggregation network and reparameterization technology to optimize information interaction between different scale feature maps, and through the Bi-Level Routing Attention (BRA) mechanism, it effectively captures critical feature information of small objects. Finally, we propose the Shape Mean Perpendicular Distance Intersection over Union (SMPDIoU) loss function. The method comprehensively considers the shape and size of detection boxes, enhances the model's focus on the attributes of detection boxes, and provides a more accurate bounding box regression loss calculation method. To demonstrate our approach's efficacy, we conducted comprehensive experiments across the RSOD, NWPU VHR-10, and VisDrone2019 datasets. The experimental results show that the HP-YOLOv8 achieves 95.11%, 93.05%, and 53.49% in the mAP@0.5 metric, and 72.03%, 65.37%, and 38.91% in the more stringent mAP@0.5:0.95 metric, respectively.

Selectively maintaining an object's feature in visual working memory: A comparison between highly discriminable and fine-grained features.

Selectively maintaining information is an essential function of visual working memory (VWM). Recent VWM studies have mainly focused on selective maintenance of objects, leaving the mechanisms of selectively maintaining an object's feature in VWM unknown. Based on the interactive model of perception and VWM, we hypothesized that there are distinct selective maintenance mechanisms for objects containing fine-grained features versus objects containing highly discriminable features. To test this hypothesis, we first required participants to memorize a dual-feature object (colored simple shapes vs. colored polygons), and informed them about the target feature via a retro-cue. Then a visual search task was added to examine the fate of the irrelevant feature. The selective maintenance of an object's feature predicted that the irrelevant feature should be removed from the active state of VWM and should not capture attention when presented as a distractor in the visual search task. We found that irrelevant simple shapes impaired performance in the visual search task (Experiment 1). However, irrelevant polygons did not affect visual search performance (Experiment 2), and this could not be explained by decay of polygons (Experiment 3) or by polygons not capturing attention (Experiment 4). These findings suggest that VWM adopts dissociable mechanisms to selectively maintain an object's feature, depending on the feature's perceptual characteristics.