Image Object Detection Research Articles

Defect Detection of Tomato Fruit Using Image Dataset

Abstract: The tomato crop is more susceptible to disease than any other vegetable, and it can be infected with over 200 diseases caused by different pathogens worldwide. Tomato plant diseases have become a challenge to food security globally. Currently, diagnosing and preventing tomato plant diseases is a challenge due to the lack of essential methods or tools.The traditional techniques of detecting plant disease are arduous and error-prone. Utilizing precise or automatic detection methods in spotting early plant disease can improve the quality of food production and reduce adverse effects. Deep learning has significantly increased the recognition accuracy of image classification and object detection systems in recent years. In this study, a 15-layer convolutional neural network is proposed as the backbone for single shot detector (SSD) to improve the detection of healthy, and three classes of tomato fruit diseases. The proposed model performance is compared with ResNet-50, AlexNet, VGG 16, and VGG19 as the backbone for Single shot detector. The findings of the experiment showed that the proposed CNN-SDD achieved 98.87% higher detection accuracy, which outperformed state-of-the-art models.

Vision-Based Asphalt Pavement Defect Detection Using Deep CNN with BIM Integration

Abstract: The major objective of the research is to predict the defects on the asphalt pavement which is the main concern for the smooth movement of vehicles by using deep CNN. CNN consists of different operations that is image classification, object detection, and image segmentation. We use detectron2 for labelling the defective part to detect the defect accurately. The data set which consists of images of defects of pavement is collected and used for training the algorithm and some of the collected images are used for the testing. The suggested approach makes use of deep learning to automatically detect and categorize a range of asphalt pavement defects, such as surface distress, potholes, and cracks. The training, validation, and testing of the model are conducted on an extensive dataset of annotated pavement photographs. The trained model shows resilience and the ability to generalize, which qualifies it for practical uses in infrastructure management and pavement upkeep. This study advances automated pavement inspection systems, which provide a practical and economical means of determining which maintenance tasks should be prioritized to maintain safer and smoother roads and this approach predicted the defects with an accuracy of 98%.

Review on Detecting Pneumonia in Deep Learning

Deep learning is a machine learning technique that has been optimized for image classification and object detection. Deep learning has brought huge advancement to the medical field as it helps to diagnose various diseases through computed tomography (CT) scan or X-ray images. Pneumonia is a respiratory disease, and it is one of the killer diseases that causes numerous death all around the world. In 2019, the outbreak of COVID-19 has increased the number of pneumonia patients tremendously. With the increasing number of patients, the clinical and medical facilities have become insufficient. The lack of doctors and radiologists to diagnose pneumonia has caused a high number of patients to be misdiagnosed. Chest image is one of the most effective methods to diagnose this disease, however, examining the X-ray or CT images requires specialists such as radiologists. Meanwhile, examining chest CT or X-ray images might be subjective as the presence of pneumonia can be unclear in the images. The main objective of this paper is to provide a comprehensive review of recent advancement in the diagnosis of pneumonia with deep learning, including state-of-art methodology, datasets, discussion, challenges, and future improvements.

Image Recognition System for Mechanical Parts Based on SSD and YOLOv5

With the development of industrial automation, the importance of mechanical parts in the production process is becoming increasingly important. In recent years, deep learning technology has made breakthrough progress in fields such as image recognition and object detection. Deep learning models can learn a large amount of data through multiple epochs, automatically extract features from it, and achieve image recognition and object detection. This provides potent technical assistance for the design of mechanical component recognition systems. In industrial production, there are a wide variety of mechanical parts with complex shapes, and the differences between different models are often very small, making it difficult to identify them manually. Therefore, it is necessary to develop a system that can automatically identify mechanical parts to improve production efficiency and quality. This article takes mechanical parts as the research object and studies the recognition technology of mechanical parts based on deep learning and image recognition. The main content of the article includes the following aspects. Research was conducted on image quality enhancement methods for images with different characteristics, and different image preprocessing methods were applied to different images. YOLOv5 and SSD (Single Shot MultiBox Detector) algorithms are used to perform deep learning and recognition detection on the processed images, respectively. By comprehensively comparing the differences between the two algorithms, YOLOv5 was selected as the focus of this design, and based on this, deep training was conducted, and the code was imported into the embedded system.

Advancements in Self-Supervised Learning for Remote Sensing Scene Classification: Present Innovations and Future Outlooks

Deep learning methodologies have significantly advanced the fields of computer vision and machine learning, enhancing performance across various tasks like classification, regression, and detection. In remote sensing for Earth observation, deep neural networks have propelled state-of-the-art results. However, a major drawback is their dependence on large annotated datasets, necessitating extensive human effort, especially in specialized domains like medical imaging or remote sensing. To mitigate this reliance on annotations, several self-supervised representation learning techniques have emerged, aiming to learn unsupervised image representations applicable to downstream tasks such as image classification, object detection, or semantic segmentation. Consequently, self-supervised learning approaches have gained traction in remote sensing. This article surveys the foundational principles of various self-supervised methods, focusing on scene classification tasks. We elucidate key contributions, analyze experimental setups, and synthesize findings from each study. Furthermore, we conduct comprehensive experiments on two public scene classification datasets to evaluate and benchmark different self-supervised models

Localization-aware logit mimicking for object detection in adverse weather conditions

Adverse weather conditions would decrease the image quality, leading to a sharp decline in detection accuracy. Most of the researches focus on object detection in fine weather conditions, rather than in adverse weather conditions. Recently, some methods attempt to reduce the gap between degraded images and clean images to improve the detection accuracy in adverse weather conditions. Specifically, these methods usually conduct image restoration and object detection in a sequential way or by joint learning. While these methods can improve detection accuracy to a certain extent, image restoration models may introduce noise or artifacts and increase computational burden, limiting the accuracy and efficiency of object detection in adverse weather conditions. In this paper, we propose a knowledge distillation-based method, Localization-aware Logit Mimicking (LaLM), to improve detection accuracy in adverse weather conditions by reducing the gap between degraded images and clean images at the prediction level, rather than the image level. Moreover, the localization quality is designed as the mimicking target to make the knowledge distillation more effective. Experiments conducted on three popular benchmarks (i.e., RTTS, ExDark, and RID) demonstrate that our LaLM can achieve the state-of-the-art detection accuracy and inference speed in foggy, rainy, and low-light conditions. Code is available at: https://github.com/VIPLab-CQU/LaLM.

Advancements in Self-Supervised Learning for Remote Sensing Scene Classification: Present Innovations and Future Outlooks

Deep learning methodologies have significantly advanced the fields of computer vision and machine learning, enhancing performance across various tasks like classification, regression, and detection. In remote sensing for Earth observation, deep neural networks have propelled state-of-the-art results. However, a major drawback is their dependence on large annotated datasets, necessitating extensive human effort, especially in specialized domains like medical imaging or remote sensing. To mitigate this reliance on annotations, several self-supervised representation learning techniques have emerged, aiming to learn unsupervised image representations applicable to downstream tasks such as image classification, object detection, or semantic segmentation. Consequently, self-supervised learning approaches have gained traction in remote sensing. This article surveys the foundational principles of various self-supervised methods, focusing on scene classification tasks. We elucidate key contributions, analyze experimental setups, and synthesize findings from each study. Furthermore, we conduct comprehensive experiments on two public scene classification datasets to evaluate and benchmark different self-supervised models

SIAM: A parameter-free, Spatial Intersection Attention Module

Attention mechanisms have been shown to play a crucial role in enhancing visual perception tasks. However, in most existing approaches, channel and spatial attention maps are estimated separately without considering the varying importance of each other. This results in a coarse attention weight for objects of interest from a holistic 3-D perspective. To address this issue, we propose a novel Parameter-free Spatial Intersection Attention Module (SIAM), which estimates 3D attention maps with spatial intersection using a parameter-free way. Specifically, SIAM first generates two independent mean queries from two spatial axes and views input as keys. Then, by computing a dot product between these mean queries and keys, SIAM generates two cross-dimension (channel and spatial) attention maps from two spatial directions and combines them into 3-D attention maps. By doing so, the produced attention maps reason important areas with spatial intersection, which can capture location-aware information to facilitate difficult objects’ location in the images. We evaluate our method in image classification, object detection, and object segmentation tasks. Extensive experimental results consistently demonstrate our approach is superior to its counterparts.

Lightweight two-stage transformer for low-light image enhancement and object detection

In low-light conditions, due to the loss of image details the visual tasks are challenging. To achieve real-time image enhancement and improve the accuracy of object detection task, we propose a lightweight two-stage Transformer. First, we use dynamic convolution to improve the adaptability of the network to different samples and preliminarily enhance the image through predicting the multiplicative and additive maps of the least squares method. In the second stage, we propose a FFT-Guidance Block (FGB) to obtain frequency components for explicit modeling, guiding the recovery of image potential information. In addition, joint our model with YOLOv3 to build a dark object detection framework, and we only use normal detection loss to simplify the training process. On the LOLv2 dataset, our model achieves advanced results. The enhanced model maintains good performance while the parameters are only 0.050M, and increase the accuracy of the downstream object detection task. The detection framework reaches 77.9% and 60.2 in mAP and FPS respectively on ExDark dataset, which can be better robust in dark conditions.

Attention-based convolution neural network for magnetic tile surface defect classification and detection

Effectively identifying surface defects in magnetic tiles has proven to be highly challenging due to limited sample availability and irrelevant background interference, which also plays a crucial role in significantly influencing the lifespan and reliability of permanent magnet motors. To address these challenges, our study draws inspiration from a comprehensive analysis of the retinal attention mechanism and proposes three guiding criteria: multi-level resolution, what to look for, and where to look at. These criteria are utilized as foundational principles to enhance the representation learning capability of designed neural network structures through the incorporation of the retinal attention mechanism. Subsequently, based on these guiding criteria, we introduce a novel convolutional retinal attention block (CRAB) to learn discriminative and robust feature representations for magnetic tile surface defect classification and detection. The proposed CRAB comprises three modules: multi-resolution module (MRM), global attention aggregation module (GAAM), and local attention aggregation module (LAAM), designed to extract discriminative and robust features by refining meaningful information and suppressing redundant ones. Comprehensive experimental results across image classification and object detection tasks demonstrate that the proposed CRAB outperforms existing methods such as SE, ECA, and CBAM, and can effectively amplify the representation power across various backbone networks, including VGG-16, GoogLeNet, ResNet-18, and ResNet-50. An evaluation on surface defect classification and detection tasks for industrial magnetic tiles further shows that CRAB achieves accuracies of 99.50% and 96.98%, respectively. These results emphasize the promising application prospects of the proposed method in detecting industrial surface defects amid expansive and inconsequential backgrounds. The code of the proposed method is available at: https://github.com/KWflyer/CRAB.

Vision Transformers with Hierarchical Attention

This paper tackles the high computational/space complexity associated with multi-head self-attention (MHSA) in vanilla vision transformers. To this end, we propose hierarchical MHSA (H-MHSA), a novel approach that computes sell-attention in a hierarchical fashion. Specifically, we first divide the input image into patches as commonly done, and each patch is viewed as a token. Then, the proposed H-MHSA learns token relationships within local patches, serving as local relationship modeling. Then, the small patches are merged into larger ones, and H-MHSA models the global dependencies for the small number of the merged tokens. At last, the local and global attentive features are aggregated to obtain features with powerful representation capacity. Since we only calculate attention for a limited number of tokens at each step, the computational load is reduced dramatically. Hence, H-MHSA can efficiently model global relationships among tokens without sacrificing fine-grained information. With the H-MHSA module incorporated, we build a family of hierarchical-attention-based transformer networks, namely HAT-Net. To demonstrate the superiority of HAT-Net in scene understanding, we conduct extensive experiments on fundamental vision tasks, including image classification, semantic segmentation, object detection and instance segmentation. Therefore, HAT-Net provides a new perspective for vision transformers. Code and pretrained models are available at https://github.com/yun-liu/HAT-Net.

Low-quality image object detection based on reinforcement learning adaptive enhancement

The task of object detection for low-quality images in multiple scenarios remains a challenging task. To alleviate this problem, a novel approach to improve the detection performance of the model on low-quality images using deep reinforcement learning is proposed. The approach differs from methods that involve data labeling and retraining, as the detection results of the model are utilized to formulate a reward function. This enables deep reinforcement learning to transform low-quality images into high-quality images through image enhancement, ultimately leading to an improvement in the model's recognition performance on low-quality images. An image enhancement tool chain (IETC) has been prepared, offering a flexible and lightweight algorithm selection for the Dueling Deep Q-Network-based Tool Selector (DDQN-TS). By adding a “Pass” option, Enable DDQN-TS to learn the relative relationship between performing and not performing image enhancement. To mitigate the potential inconsistency between the output of the detection model and the actual situation caused by negative samples, which can lead to incorrect decisions made by DDQN-TS, the introduction of a “thresh” parameter regulates the level of intervention in target detection by DIE. The experimental results demonstrate the effectiveness and portability of the DIE model in environments with noise, fog, and uneven brightness.

Investigating attention mechanisms for plant disease identification in challenging environments

There is an increasing demand for efficient and precise plant disease detection methods that can quickly identify disease outbreaks. For this, researchers have developed various machine learning and image processing techniques. However, real-field images present challenges due to complex backgrounds, similarities between different disease symptoms, and the need to detect multiple diseases simultaneously. These obstacles hinder the development of a reliable classification model. The attention mechanisms emerge as a critical factor in enhancing the robustness of classification models by selectively focusing on relevant regions or features within infected regions in an image. This paper provides details about various types of attention mechanisms and explores the utilization of these techniques for the machine learning solutions created by researchers for image segmentation, feature extraction, object detection, and classification for efficient plant disease identification. Experiments are conducted on three models: MobileNetV2, EfficientNetV2, and ShuffleNetV2, to assess the effectiveness of attention modules. For this, Squeeze and Excitation layers, the Convolutional Block Attention Module, and transformer modules have been integrated into these models, and their performance has been evaluated using different metrics. The outcomes show that adding attention modules enhances the original models' functionality.

Discretization and decoupled knowledge distillation for arbitrary oriented object detection

In recent years, lightweight object detection networks have been increasingly applied to remote sensing platforms due to their fast inference speed and flexible deployment advantages. Knowledge distillation methods have been widely used to reduce the performance gap between large and small models, and many studies have combined knowledge distillation with object detection. However, existing knowledge distillation methods often overlook the transfer of localization knowledge. Therefore, this paper proposes a method called Discretized Position Knowledge Distillation (DPKD) to improve the use of knowledge distillation in object detection. Specifically, the DPKD method incorporates a Discretization Algorithm Module (DAM), which leverages both general probability distribution and cross-Gaussian distribution to transfer high-quality bounding box position and pose information. Additionally, the Position Knowledge Distillation (PKD) method splits the target and non-target bounding boxes to form the distillation loss function, addressing the issue of missing background knowledge transfer during the distillation process. To further enhance the learning of high-quality bounding boxes, a Region Weighting Module (RWM) based on EIoU is introduced in DPKD, assigning weights to the various bounding boxes in the teacher's output. The effectiveness of DPKD in the field of remote sensing image object detection in multi-modal scenarios was verified through multi-modal training on the publicly available DOTA and HRSID datasets.

An Approach towards to Real Time Face Detection using Haar Cascade Algorithm and Machine Learning Technique

Human visual awareness is currently a trending subject within the realm of machine vision. In domains like video monitoring, human-machine interaction, facial identification, and image organization, the initial step often involves locating and recognizing human faces. However, this task can prove challenging. Despite the availability of generic facial images, the processes of facial recognition and expression analysis are necessary. The exploration of computer-based recognition of unbiased facial data remains a relatively uncharted area of study. Facial recognition stands out as one of the remarkable achievements of AI research, captivating the interest of many tech enthusiasts eager to comprehend its underlying mechanisms. Let's delve deeper into this topic to gain insight into its functioning. This study elucidates how deep learning, a crucial component of computer science, can be employed to detect faces using various libraries within OpenCV and Python. The article will propose a cutting-edge technology for real-time human face detection, applicable across a spectrum of platforms including computers, smartphones, and software applications. This approach proves both efficient and effective in detecting faces within images. Furthermore, the article delves into popular OpenCV applications and classifiers utilized in tasks such as image manipulation, facial identification, object detection, and facial recognition. Lastly, various literary assessments are discussed, focusing on applications within computer vision disciplines like facial detection and recognition, as well as identifying emotions such as sorrow, anger, and joy, along with determining gender and age

Deep learning techniques to detect rail indications from ultrasonic data for automated rail monitoring and maintenance

The increasing number of passengers and services using railways and the corresponding increase in rail use has caused the acceleration of rail wear and surface defects which makes rail defect identification an important issue for rail maintenance and monitoring to ensure safe and efficient operation. Traditional visual inspection methods for identifying rail defects are time-consuming, less accurate, and associated with human errors. Deep learning has been used to improve railway maintenance and monitoring tasks. This study aims to develop a structured model for detecting railway artifacts and defects by comparing different deep-learning models using ultrasonic image data. This research showed whether it is practical to identify rail indications using image classification and object detection techniques from ultrasonic data and which model performs better among the above-mentioned methods. The methodology includes data processing, labeling, and using different conventional neural networks to develop the model for both image classification and object detection. The results of CNNs for image classification, and YOLOv5 for object detection show 98%, and 99% accuracy respectively. These models can identify rail artifacts efficiently and accurately in real-life scenarios, which can improve automated railway infrastructure monitoring and maintenance.

AFDet: alignment and focusing for aerial object detection

ABSTRACT Object detection in aerial images has become a focus in recent years due to the expansion of its application fields. Aerial objects have the characteristics of multi-scale, arbitrary angle and dense arrangement, which brings considerable challenges to the task. Based on the efficient one-stage detectors, most methods design modules to adaptively align the feature receptive field with the anchors to improve the accuracy of the bounding box regression and calculate loss according to parameter bias. Current methods are not adequately data-driven in feature optimization and the angle parameter regression faces boundary problems. To handle these problems, in this letter, we propose an Align-Focus detector (AFDet) for aerial image object detection. We introduce the ideology of deformable multi-head self-attention to feature optimization and design a new Align-Focus Module (AFM) which can sensitively focus feature encoding response to the real texture area of the objects. In addition, we apply KF-IOU Loss to solve the boundary problem. We conduct necessary experiments on DOTA and HRSC2016 datasets, and AFDet achieves the highest mAP performance compared to the existing methods.

Optimizing DNN training with pipeline model parallelism for enhanced performance in embedded systems

Deep Neural Networks (DNNs) have gained widespread popularity in different domain applications due to their dominant performance. Despite the prevalence of massively parallel multi-core processor architectures, adopting large DNN models in embedded systems remains challenging, as most embedded applications are designed with single-core processors in mind. This limits DNN adoption in embedded systems due to inefficient leveraging of model parallelization and workload partitioning. Prior solutions attempt to address these challenges using data and model parallelism. However, they lack in finding optimal DNN model partitions and distributing them efficiently to achieve improved performance.This paper proposes a DNN model parallelism framework to accelerate model training by finding the optimal number of model partitions and resource provisions. The proposed framework combines data and model parallelism techniques to optimize the parallel processing of DNNs for embedded applications. In addition, it implements the pipeline execution of the partitioned models and integrates a task controller to manage the computing resources. The experimental results for image object detection demonstrate the applicability of our proposed framework in estimating the latest execution time and reducing overall model training time by almost 44.87% compared to the baseline AlexNet convolutional neural network (CNN) model.

DIGCN: A Dynamic Interaction Graph Convolutional Network Based on Learnable Proposals for Object Detection

We propose a Dynamic Interaction Graph Convolutional Network (DIGCN), an image object detection method based on learnable proposals and GCN. Existing object detection methods usually work on dense candidates, resulting in redundant and near-duplicate results. Meanwhile, non-maximum suppression post-processing operations are required to eliminate negative effects, which increases the computational complexity. Although the existing sparse detector avoids cumbersome post-processing operations, it ignores the potential relationship between objects and proposals, which hinders detection accuracy improvement. Therefore, we propose a dynamic interaction GCN module in the DIGCN, which performs dynamic interaction and relational modeling on the proposal boxes and proposal features to improve the object detection accuracy. In addition, we introduce a learnable proposal method with a sparse set of learned object proposals to eliminate a huge number of hand-designed object candidates, avoiding complicated tasks such as object candidate design and many-to-one label assignment, and reducing object detection model complexity to a certain extent. DIGCN demonstrates accuracy and run-time performance on par with the well-established and highly optimized detector baselines on the challenging COCO dataset, e.g. with the ResNet-101FPN as the backbone our method attains the accuracy of 46.5 AP while processing 13 frames per second. Our work provides a new method for object detection research.

Application of convolutional neural networks trained on optical images for object detection in radar images

Due to the small number of annotated radar image datasets, the use of optical images for training neural networks designed to detect objects in radar images seems promising. However, optical images have some significant differences from radar images and an experimental investigation of this possibility is required. In this work we investigate the applicability of such an approach and show that in the case of detection of ships good results can be achieved. In addition, it is shown that preliminary filtering of speckle noise can improve the results.