Object Recognition Research Articles

Zero-shot counting with a dual-stream neural network model.

To understand a visual scene, observers need to both recognize objects and encode relational structure. For example, a scene comprising three apples requires the observer to encode concepts of "apple" and "three." In the primate brain, these functions rely on dual (ventral and dorsal) processing streams. Object recognition in primates has been successfully modeled with deep neural networks, but how scene structure (including numerosity) is encoded remains poorly understood. Here, we built a deep learning model, based on the dual-stream architecture of the primate brain, which is able to count items "zero-shot"-even if the objects themselves are unfamiliar. Our dual-stream network forms spatial response fields and lognormal number codes that resemble those observed in the macaque posterior parietal cortex. The dual-stream network also makes successful predictions about human counting behavior. Our results provide evidence for an enactive theory of the role of the posterior parietal cortex in visual scene understanding.

AI-Powered Disinfection Force: Harnessing 5G for Intelligent Distributed Object Recognition and Sanitation Fleet

With the increasing prevalence of embedded cameras in devices such as phones, robots, and vehicles, there’s a growing need to enhance their perception capabilities through advanced deep learning techniques. These systems, however, often struggle under suboptimal conditions like environmental noise. The paper addresses these challenges by proposing a 5G-enabled solution that uses a fleet of unmanned ground vehicles (UGVs) for coordinated disinfection tasks in hospital environments. Utilizing high-bandwidth, low-latency 5G networks, these UGVs can share high-quality images for distributed object recognition tasks and execute precise disinfection routines using onboard UV lamps. The proposed system includes an optimal formation algorithm for UGVs to maintain effective positioning and coordination. By leveraging 5G connectivity, the fleet can efficiently exchange sensor data and perform real-time, innovative object recognition and disinfection, presenting a significant advancement in intelligent sanitation technology.

Vision assistant for visually impaired individuals

Visually impaired individuals face significant challenges in accessing environmental information, social interactions, and safety in their daily lives. This article aims to increase the independence of visually impaired individuals and simplify their daily lives using an AI-supported image and sound recognition system. The developed prototype is a pair of glasses consisting of a Raspberry Pi 4 microprocessor, integrated headphones, touch controls, and a powerful battery management system. The electronic diagram includes the integration of a camera and microprocessor. The AI algorithms developed successfully perform tasks such as object recognition, text reading, and facial recognition. The article provides a comprehensive literature review and introduces the prototype to visually impaired users. Based on the feedback from the introduction, the prototype's ease of use and effectiveness were evaluated, and improvements were made based on user feedback. The algorithms were tested for accuracy, with the prototype achieving a 78% accuracy score, close to GPT-4's 82%. Additionally, tests with 20 users resulted in an 85% user satisfaction rate.

MSSA: Multi-Representation Semantics-Augmented Set Abstraction for 3D Object Detection

Accurate recognition and localization of 3D objects is a fundamental research problem in 3D computer vision. Benefiting from transformation-free point cloud processing and flexible receptive fields, point-based methods have become accurate in 3D point cloud modeling, but still fall behind voxel-based competitors in 3D detection. We observe that the set abstraction module, commonly utilized by point-based methods for downsampling points, tends to retain excessive irrelevant background information, thus hindering the effective learning of features for object detection tasks. To address this issue, we propose MSSA, a Multi-representation Semantics-augmented Set Abstraction for 3D object detection. Specifically, we first design a backbone network to encode different representation features of point clouds, which extracts point-wise features through PointNet to preserve fine-grained geometric structure features, and adopts VoxelNet to extract voxel features and BEV features to enhance the semantic features of key points. Second, to efficiently fuse different representation features of keypoints, we propose a Point feature-guided Voxel feature and BEV feature fusion (PVB-Fusion) module to adaptively fuse multi-representation features and remove noise. At last, a novel Multi-representation Semantic-guided Farthest Point Sampling (MS-FPS) algorithm is designed to help set abstraction modules progressively downsample point clouds, thereby improving instance recall and detection performance with more important foreground points. We evaluate MSSA on the widely used KITTI dataset and the more challenging nuScenes dataset. Experimental results show that compared to PointRCNN, our method improves the AP of “moderate” level for three classes of objects by 7.02%, 6.76%, and 5.44%, respectively. Compared to the advanced point-voxel-based method PV-RCNN, our method improves the AP of “moderate” level by 1.23%, 2.84%, and 0.55% for the three classes, respectively.

Development of the intelligent traffic light system based on image processing and fuzzy control techniques

In Viet Nam's current traffic conditions, congestion and jams—especially at intersections during peak hours—present major challenges. Traditional traffic light systems, which rely on fixed timing principles, often fail to manage traffic flow efficiently, particularly when vehicle density varies significantly across different directions. This research aims to develop an intelligent traffic light system where the signal timings automatically adjust based on the vehicle density at intersections. The study uses an object recognition algorithm to identify, classify, and count vehicles. The data was then fed into a fuzzy logic model to calculate the optimal signal timings. Experimental results demonstrate an accuracy of approximately 88% in vehicle detection. The fuzzy logic model and the programmable logic controller were able to effectively compute reasonable signal timings based on real-time vehicle density. Future developments include expanding the system's functionalities, creating a user-friendly interface, and developing a management application for mobile devices.

Voucher Specimens in Taxonomy and Simpson’s Hypodigm

The problem of specifying representative specimens to serve as vouchers or ground truth specimens for species is viewed from a perspective of object recognition based on training using exemplars recognized by personal perception. In taxonomy, an ‘exemplar’ mirrors the hypodigm concept of Simpson, which refers to certain specimens being unequivocal members of a species. His concept has been discarded in most taxonomies because he did not provide procedures that distinguished it from ‘material’ or ‘sample’. However, his underlying view of a morphospecies was of a group united by shared character resemblances, probabilistically related. On this basis, the hypodigm of a taxonomic species is here restricted to specimens demonstrated as likely belonging to one population. To raise objectivity in this task, personal perception should be supplemented by morphometric analyses, several of which are demonstrated using living and Holocene samples of the oceanic zooplankter Truncorotalia crassaformis (Galloway and Wissler) from the Atlantic and Caribbean Oceans.

Recognizing objects in complex scenes: A Recent Systematic Review

This systematic review provides a comprehensive examination of recent advancements in object recognition within complex scenes, focusing on addressing challenges such as clutter, occlusion, and diverse environmental conditions. Leveraging the PRISMA framework, the study meticulously analysed a diverse range of literature from esteemed sources, employing advanced search methods on Scopus and WoS databases to identify and analyse primary research studies (n = 25). The review encompasses three key themes: Theme 1 concentrates on Image Noise Identification and Removal Techniques, while Theme 2 delves into Image Classification and Recognition under Noise. Additionally, Theme 3 explores Innovative Models and Approaches for Noise-Robust Image Analysis. Despite the progress achieved, contemporary recognition systems struggle with real-world complexities such as varied scales, lighting variations, and different viewpoints. The synthesis of findings emphasizes the necessity for innovative strategies that capitalize on contextual cues and harness the potential of deep learning to enhance precision in object recognition within intricate visual environments. The insights gleaned from this synthesis are poised to guide future research directions, informing the development of more resilient algorithms capable of navigating challenges and catalysing advancements in the field of object recognition.

Retraction Note: An agent architecture for autonomous UAV flight control in object classification and recognition missions

Retraction Note: An agent architecture for autonomous UAV flight control in object classification and recognition missions

An Energy-Efficient Dynamic Feedback Image Signal Processor for Three-Dimensional Time-of-Flight Sensors.

With the recent prominence of artificial intelligence (AI) technology, various research outcomes and applications in the field of image recognition and processing utilizing AI have been continuously emerging. In particular, the domain of object recognition using 3D time-of-flight (ToF) sensors has been actively researched, often in conjunction with augmented reality (AR) and virtual reality (VR). However, for more precise analysis, high-quality images are required, necessitating significantly larger parameters and computations. These requirements can pose challenges, especially in developing AR and VR technologies for low-power portable devices. Therefore, we propose a dynamic feedback configuration image signal processor (ISP) for 3D ToF sensors. The ISP achieves both accuracy and energy efficiency through dynamic feedback. The proposed ISP employs dynamic area extraction to perform computations and post-processing only for pixels within the valid area used by the application in each frame. Additionally, it uses dynamic resolution to determine and apply the appropriate resolution for each frame. This approach enhances energy efficiency by avoiding the processing of all sensor data while maintaining or surpassing accuracy levels. Furthermore, These functionalities are designed for hardware-efficient implementation, improving processing speed and minimizing power consumption. The results show a maximum performance of 178 fps and a high energy efficiency of up to 123.15 fps/W. When connected to the hand pose estimation (HPE) accelerator, it demonstrates an average mean squared error (MSE) of 10.03 mm, surpassing the baseline ISP value of 20.25 mm. Therefore, the proposed ISP can be effectively utilized in low-power, small form-factor devices.

Vitamin K2 Ameliorates Diabetes-Associated Cognitive Decline by Reducing Oxidative Stress and Neuroinflammation.

Diabetes, a chronic metabolic disease, affects approximately 422million people and leads to 1.5million deaths every year, It is found that 45% of individuals with diabetes eventually develop cognitive impairment. Here we study effects of Vitamin K2 on diabetes-associated cognitive decline (DACD) and its underlying mechanism. Diabetes was induced in adult Swiss albino mice with high-fat diet and a low dose (35mg/kg) of streptozotocin and measured by fasting glucose and HbA1c levels. After one week of development of diabetes, one group of animals received Vitamin K2 (100µg/kg) via oral gavage for 21 days. Then different behavioural studies, including the elevated plus maze, Morris water maze, passive avoidance test and novel object recognition test were performed followed by biochemical tests including AchE, different oxidative stress parameters (SOD, GSH, MDA, catalase, SIRT1, NRF2), inflammatory markers (TNFα, IL1β, MCP1, NFκB), apoptosis marker (Caspase 3). Hippocampal neuronal density was measured using histopathology. Vitamin K2 treatment in diabetic animals led to reduced fasting glucose and HbA1c, It could partially reverse DACD as shown by behavioural studies. Vitamin K2 adminstration reduced corticohippocampal AchE level and neuroinflammation (TNFα, IL1β, MCP1, NFκB, SIRT1). It reduced oxidative stress by increasing antioxidant enzymes (SOD, GSH, catalase), transcription factor NRF2 while reducing caspase 3. This eventually increased CA1 and CA3 neuronal density in diabetic animals. Vitamin K2 partially reverses DACD by increasing ACh while reducing the oxidative stress via Nrf2/ARE pathway and neuroinflammation, thus protecting the hippocampal neurons from diabetes associated damage.

Advancing Incremental Few-Shot Video Action Recognition with Cluster Compression and Generative Separation

Few-Shot Class Incremental Learning (FSCIL) is a trending topic in deep learning, addressing the need for models to incrementally learn novel classes, particularly in real-world scenarios where continuously emerging classes come with limited labeled samples. However, the majority of FSCIL research has been dedicated to image classification and object recognition tasks, with limited attention given to video action classification. In this paper, we present a new Cluster Compression and Generative Separation (CCGS) method for Incremental Few-Shot Video Action Recognition (iFSVAR), which introduces contrastive learning to boost the degree of class separation in the base session. Simultaneously, it creates numerous fine-grained classes with diverse semantics, effectively filling the unallocated representation space. Experimental results on UCF101, Kinetics, and Something-Something-V2[Formula: see text]demonstrate the effectiveness of the framework.

Retraction Note: Imperceptible adversarial attacks against traffic scene recognition

Retraction Note: Imperceptible adversarial attacks against traffic scene recognition

Development of a deep learning-based object recognition system for pre-stage separation to improve the recycling rate of major general-purpose plastics

ABSTRACT The generation of plastic waste has been increasing annually, necessitating various recycling methods. Among these, material recycling with low carbon emissions should be prioritized. This study aims to enhance the material recycling rate by developing a separation system using deep learning-based object recognition. To improve labeling efficiency, image data were acquired for each material type, and the performance of different labeling methods was evaluated. The average precision (AP) values for the single-material learning model using hybrid labeling (manual + automatic) were 0.947 for PET, 0.951 for PP, 0.892 for PE, and 0.896 for PS, demonstrating superior performance compared to other methods. An integrated learning model was also developed for composite materials, achieving AP values of 0.972 for PET, 0.976 for PP, 0.963 for PE, and 0.961 for PS. These results demonstrate the model’s strong applicability for plastic waste recognition.

Charting the growth through intelligence: A SWOC analysis on AI-assisted radiologic bone age estimation.

Bone age estimation (BAE) is based on skeletal maturity and degenerative process of the skeleton. The clinical importance of BAE is in understanding the pediatric and growth-related disorders; whereas medicolegally it is important in determining criminal responsibility and establishing identification. Artificial Intelligence (AI) has been used in the field of the field of medicine and specifically in diagnostics using medical images. AI can greatly benefit the BAE techniques by decreasing the intra observer and inter observer variability as well as by reducing the analytical time. The AI techniques rely on object identification, feature extraction and segregation. Bone age assessment is the classical example where the concepts of AI such as object recognition and segregation can be used effectively. The paper describes various AI based algorithms developed for the purpose of radiologic BAE and the performances of the models. In the current paper we have also carried out qualitative analysis using Strength, Weakness, Opportunities and Challenges (SWOC) to examine critical factors that contribute to the application of AI in BAE. To best of our knowledge, the SWOC analysis is being carried out for the first time to assess the applicability of AI in BAE. Based on the SWOC analysis we have provided strategies for successful implementation of AI in BAE in forensic and medicolegal context.

Mechanism of Gastrodin against neurotoxicity based on network pharmacology, molecular docking and experimental verification

BackgroundDisorders of glutamate metabolism and excessive release participat in multiple neuronal pathologies including ischemic stroke (IS), Alzheimer's disease (AD), or Parkinson's disease (PD). Recently, herbal medicines have been widely used and have shown satisfactory results in the treatment of neurological disorders. Gastrodin is a traditional Chinese medicine (TCM) used for the treatment of nerve injuries, spinal cord injuries, and some central nervous system diseases as well. This research examines the neuroprotective effects of Gastrodin against glutamate-induced neurotoxicity in neuronal cells. MethodsThe HERB database was used to explore the active ingredients and target genes of Gastrodia Elata. The STRING database and Cytoscape software were used to screen and construct the Protein-Protein Interaction (PPI). Furthermore, we used molecular docking to predict the potential targets of Gastrodin. The effects of Gastrodin were revealed by western blot, calcium imaging, membrane clamp, CCK8 and flow cytometry. Neuronal oxidative stress and damage were assessed by measuring malondialdehyde (MDA) levels and superoxide dismutase (SOD) activity. Neuronal morphology was examined using Golgi-Cox staining. Finally, animal behavior was examined using novel object recognition and fear conditioning tests. ResultsWe have obtained 22 components such as TM10, TM17, TM25 (Gastrodin), and 281 targets such as AKT, EGFR, and CDK1 through network pharmacology. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses revealed these genes were significantly enriched in protein phosphorylation, protein serine/threonine/tyrosine kinase activity, apoptosis and HIF-1 signaling pathways, etc. A higher affinity between Gastrodin and AKT was revealed by PPI analysis and molecular docking. Further, Gastrodin significantly inhibited Ca2+ influxes and excitatory synaptic transmission in cortical neurons. In addition, Gastrodin effectively alleviated neuron apoptosis, oxidative stress and damage. ConclusionGastrodin has neuroprotective effects against glutamate-induced neurotoxicity.

6-Methoxyflavone improves anxiety, depression, and memory by increasing monoamines in mice brain: HPLC analysis and in silico studies

Abstract 6-Methoxyflavone (6-MOF) is a flavonoid that has been reported to be a GABA-A receptor agonist and reverses cisplatin-induced hyperalgesia and allodynia. Considering the varied neuropharmacological profile of 6-MOF, this study was intended to determine the pharmacological effects of 6-MOF on locomotion, anxiety, novel object recognition (NOR), depression, spatial memory, socialization behavior, nest-building behavior, and depression in various groups of mice. Selected groups of mice were injected with 25, 50, and 75 mg/kg 6-MOF. Using HPLC-UV, the frontal cortex, striatum, and hippocampus of the sacrificed mice were analyzed for the levels of vitamin C, dopamine, serotonin, noradrenaline, adenosine, and its metabolites. Statistical analysis showed significant results in socialization behavior and elevated plus maze with 75 mg/kg. In Y-maze, NOR 6-MOF showed significant results at all three doses, while in tail suspension test (TST), 50 and 75 mg/kg showed significant results; however, no statistical significance was observed in nest-building behavior; 50 and 75 mg/kg 6-MOF showed significant results in the Morris water maze. 6-MOF raised vitamin C levels in the frontal cortex and hippocampus. Serotonin, dopamine, and nor-adrenaline levels were raised in the hippocampus and striatum. It has also imparted region-specific neuroprotection by improving adenosine and its metabolite levels. In silico studies performed using PyRx have shown that the minimum binding energy of 6-MOF with antioxidant enzyme is −7.1 k/cal/mol. The binding energy showed that 6-MOF was successfully docked with an anti-oxidant enzyme. In conclusion, in silico and behavioral studies showed that 6-MOF can be a potential candidate for the treatment of cognitive decline, anxiety, and depression.

Comparative analysis of the impact of repetitive transcranial magnetic stimulation and burst transcranial magnetic stimulation at different frequencies on memory function and neuronal excitability of mice

Transcranial magnetic stimulation (TMS) as a non-invasive neuroregulatory technique has been applied in the clinical treatment of neurological and psychiatric diseases. However, the stimulation effects and neural regulatory mechanisms of TMS with different frequencies and modes are not yet clear. This article explores the effects of different frequency repetitive transcranial magnetic stimulation (rTMS) and burst transcranial magnetic stimulation (bTMS) on memory function and neuronal excitability in mice from the perspective of neuroelectrophysiology. In this experiment, 42 Kunming mice aged 8 weeks were randomly divided into pseudo stimulation group and stimulation groups. The stimulation group included rTMS stimulation groups with different frequencies (1, 5, 10 Hz), and bTMS stimulation groups with different frequencies (1, 5, 10 Hz). Among them, the stimulation group received continuous stimulation for 14 days. After the stimulation, the mice underwent new object recognition and platform jumping experiment to test their memory ability. Subsequently, brain slice patch clamp experiment was conducted to analyze the excitability of granulosa cells in the dentate gyrus (DG) of mice. The results showed that compared with the pseudo stimulation group, high-frequency (5, 10 Hz) rTMS and bTMS could improve the memory ability and neuronal excitability of mice, while low-frequency (1 Hz) rTMS and bTMS have no significant effect. For the two stimulation modes at the same frequency, their effects on memory function and neuronal excitability of mice have no significant difference. The results of this study suggest that high-frequency TMS can improve memory function in mice by increasing the excitability of hippocampal DG granule neurons. This article provides experimental and theoretical basis for the mechanism research and clinical application of TMS in improving cognitive function.

Acute Administration of Edaravone Improves Cognitive Impairment in a Mouse Model of mPFC Ischemia: Crosstalk Between Necroptosis, Neuroinflammation, and Antioxidant Defense.

Edaravone (Eda), a well-known free radical scavenger, has been reported as a possible therapeutic agent for ischemic stroke patients' recovery. This study aimed to investigate the effects of time-dependent treatment with Eda on medial prefrontal cortex (mPFC) ischemia. Mice were randomly allocated into six groups: control, sham, normal saline, Eda-I, Eda-II, and Eda-III. After induction of a photothrombotic ischemia in the mPFC region, Eda-I, Eda-II, and Eda-III groups received 3mg/kg Eda intraperitoneally at the times of 0, 2, and 6h post-surgery. After 1day of recovery, the mice underwent behavioral tests (open field, novel object recognition, and T-maze). Next, necroptosis, NOD-like receptor protein 3 (NLRP3), and nuclear factor erythroid 2-related factor 2 (Nrf2) pathway-related protein levels were measured in the lesioned area using western blot analysis. For double confirmation, IL-1β and IL-18 were also assessed by immunofluorescence in the area. Further, histological evaluations were performed to measure tissue damage. The results showed that mPFC ischemia impaired recognition and spatial working memory without affecting locomotor activity, while immediate Eda administration improved cognitive impairments. Furthermore, acute Eda treatment reduced RIP1, RIP3, and MLKL levels, inhibited NLRP3 inflammasome proteins (NLRP3, ASC, and Cas1), decreased IL-1β and IL-18, upregulated Nrf2 and its targets (NQO-1 and HO-1), and diminished tissue damage. Our results highlighted the effects of acute administration of Eda post-stroke on improving cognitive impairments by suppressing necroptosis and NLRP3 inflammasome pathways and activating the Nrf2 antioxidant defense mechanism.

Evaluating Convolutional Neural Network Architecture for Historical Topographic Hardcopy Maps Analysis: A Study on Training and Validation Accuracy Variation

Convolutional Neural Networks (CNN) are widely used for image analysis tasks, including object detection, segmentation, and recognition. Given the advanced capability, this study evaluates the effectiveness and performance of CNN architecture for analysing Historical Topographic Hardcopy Maps (HTHM) by assessing variations in training and validation accuracy. The lack of research specifically dedicated to CNN’s application in analysing topographic hardcopy maps presents an opportunity to explore and address the unique challenges associated with this domain. While existing studies have predominantly focused on satellite imagery, this study aims to uncover valuable insights, patterns, and characteristics inherent to HTHM through customised CNN approaches. This study utilises a standard CNN architecture and tests the model’s performance with different epoch settings (20, 40, and 60) using varying dataset sizes (288, 636, 1144, and 1716 images). The results indicate that the optimal operation point for training and validation accuracy is achieved at epoch 40. Beyond epoch 40, the widening gap between training and validation accuracy suggests overfitting. Hence, adding more epochs does not significantly improve accuracy beyond the optimum phase. The experiment also shows that the CNN model obtains a training accuracy of 98%, validation accuracy of 67%, and F1-score overall performance of 77%. The analysis demonstrates that the CNN model performs reasonably well in classifying instances from the HTHM dataset. These findings contribute to a better understanding of the strengths and limitations of the model, providing valuable insights for future research and refinement of classification approaches in the context of topographic hardcopy map analysis.

Towards a Taxonomy Machine: A Training Set of 5.6 Million Arthropod Images

The taxonomic identification of organisms from images is an active research area within the machine learning community. Current algorithms are very effective for object recognition and discrimination, but they require extensive training datasets to generate reliable assignments. This study releases 5.6 million images with representatives from 10 arthropod classes and 26 insect orders. All images were taken using a Keyence VHX-7000 Digital Microscope system with an automatic stage to permit high-resolution (4K) microphotography. Providing phenotypic data for 324,000 species derived from 48 countries, this release represents, by far, the largest dataset of standardized arthropod images. As such, this dataset is well suited for testing the efficacy of machine learning algorithms for identifying specimens into higher taxonomic categories.