Object Detection Research Articles

Refined feature enhancement network for object detection

Convolutional neural networks-based object detection techniques have achieved positive performances. However, due to the limitations of local receptive field, some existing object detection methods cannot effectively capture global information in feature extraction phases, and thus lead to unsatisfactory detection performance. Moreover, the feature information extracted by the backbone network may be redundant. To alleviate these problems, in this paper we propose a refined feature enhancement network (RFENet) for object detection. Specifically, we first propose a feature enhancement module (FEM) to capture more global and local information from feature maps with certain long-range dependencies. We further propose a multi-branch dilated attention mechanism (MDAM) to refine the extracted features in a weighted form, which can select more important spatial and channel information and broaden the receptive field of the network. Finally, we validate RFENet on MS-COCO2017, PASCAL VOC2012, and PASCAL VOC07+12 datasets, respectively. Compared to the baseline network, our RFENet improves by 2.4 AP on MS-COCO2017 dataset, 3.4 mAP on PASCAL VOC2012 dataset, and 2.7 mAP on PASCAL VOC07+12 dataset. Extensive experiments show that our RFENet can perform competitively on different datasets. The code is available at https://github.com/object9detection/RFENet.

Clinical efficacy of DSA-based features in predicting outcomes of acupuncture intervention on upper limb dysfunction following ischemic stroke

Background and objectivesThis study aimed to employ machine learning techniques to predict the clinical efficacy of acupuncture as an intervention for patients with upper limb motor dysfunction following ischemic stroke, as well as to assess its potential utility in clinical practice.MethodsMedical records and digital subtraction angiography (DSA) imaging data were collected from 735 ischemic stroke patients with upper limb motor dysfunction who were treated with standardized acupuncture at two hospitals. Following the initial screening, 314 patient datasets that met the inclusion criteria were selected. We applied three deep-learning algorithms (YOLOX, FasterRCNN, and TOOD) to develop the object detection model. Object detection results pertaining to the cerebral vessels were integrated into a clinical efficacy prediction model (random forest). This model aimed to classify patient responses to acupuncture treatment. Finally, the accuracies and discriminative capabilities of the prediction models were assessed.ResultsThe object detection model achieved an optimal recognition rate, The mean average precisions of YOLOX, TOOD, and FasterRCNN were 0.61, 0.7, and 0.68, respectively. The prediction accuracy of the clinical efficacy model reached 93.6%, with all three-treatment response classification area under the curves (AUCs) exceeding 0.95. Feature extraction using the prediction model highlighted the significant influence of various cerebral vascular stenosis sites within the internal carotid artery (ICA) on prediction outcomes. Specifically, the initial and C1 segments of the ICA had the highest predictive weights among all stenosis sites. Additionally, stenosis of the middle cerebral, anterior cerebral, and posterior cerebral arteries exerted a notable influence on the predictions. In contrast, the stenosis sites within the vertebral artery exhibited minimal impact on the model's predictive abilities.ConclusionsResults underscore the substantial predictive influence of each cerebral vascular stenosis site within the ICA, with the initial and C1 segments being pivotal predictors.

A Refined and Efficient CNN Algorithm for Remote Sensing Object Detection

Remote sensing object detection (RSOD) plays a crucial role in resource utilization, geological disaster risk assessment and urban planning. Deep learning-based object-detection algorithms have proven effective in remote sensing image studies. However, accurate detection of objects with small size, dense distribution and complex object arrangement remains a significant challenge in the remote sensing field. To address this, a refined and efficient object-detection algorithm (RE-YOLO) has been proposed in this paper for remote sensing images. Initially, a refined and efficient module (REM) was designed to balance computational complexity and feature-extraction capabilities, which serves as a key component of the RE_CSP block. RE_CSP block efficiently extracts multi-scale information, overcoming challenges posed by complex backgrounds. Moreover, the spatial extracted attention module (SEAM) has been proposed in the bottleneck of backbone to promote representative feature learning and enhance the semantic information capture. In addition, a three-branch path aggregation network (TBPAN) has been constructed as the neck network, which facilitates comprehensive fusion of shallow positional information and deep semantic information across different channels, enabling the network with a robust ability to capture contextual information. Extensive experiments conducted on two large-scale remote sensing datasets, DOTA-v1.0 and SCERL, demonstrate that the proposed RE-YOLO outperforms state-of-the-art other object-detection approaches and exhibits a significant improvement in generalization ability.

Nets4Learning: A Web Platform for Designing and Testing ANN/DNN Models

Nowadays, any research discipline is interested in tackling its problems with artificial intelligence and, therefore, is demanding knowledge and frameworks with the aim of developing and using intelligent methods. Within this scenario, neural networks stand out for the important results they have achieved. This paper introduces Nets4Learning, a web platform for designing, training and testing artificial/deep neural network models. The application deals with some of the most popular tasks in the data science field such as tabular classification, regression, image classification and object detection. Nets4Learning has been designed so that any researcher from any discipline can easily develop neural network models without special programming or digital skills. In fact, the user does not have to install anything as the application is publicly available and can be accessed from any device. The site also has manuals, glossaries, etc., and all this code is available on GitHub.

YOLO vs CNN : A Comparative Study on Object Detection

Abstract—With the availability of large amount of data and increasing demand of digitizing the visual systems, lot of research on object detection techniques has been the foremost requirement since last ten years. As the technological advancements are pacing at such a higher rate with increasing computational power and new architectures for CNN has already been proposed in a successive manner to enhance the detection techniques. For Real time Object identification, CNN has proposed several architectures but YOLO proves itself better than others based on the speed of identification. This paper presents the brief overview and comparative analysis of various object detection techniques like SSD, RCNN, YOLO and their subsequent versions for different applications. On comparing, it is found that single stage detectors outperform two stage detectors. Keywords—Pascal VOC, MS COCO, YOLO, CNN, Object detection, Object Recognition, SSD.

A Black-Box Watermarking Modulation for Object Detection Models

Training a Deep Neural Network (DNN) from scratch comes with a substantial cost in terms of money, energy, data, and hardware. When such models are misused or redistributed without authorisation, the owner faces significant financial and intellectual property (IP) losses. Therefore, there is a pressing need to protect the IP of Machine Learning models to avoid these issues. ML watermarking emerges as a promising solution for model traceability. Watermarking has been well-studied for image classification models, but there is a significant research gap in its application to other tasks like object detection, for which no effective methods have been proposed yet. In this paper, we introduce a novel black-box watermarking method for object detection models. Our contributions include a watermarking technique that maps visual information to text semantics and a comparative study of fine-tuning techniques’ impact on watermark detectability. We present the model’s detection performance and evaluate fine-tuning strategies’ effectiveness in preserving watermark integrity.

Near real‐time monitoring of wading birds using uncrewed aircraft systems and computer vision

Wildlife population monitoring over large geographic areas is increasingly feasible due to developments in aerial survey methods coupled with the use of computer vision models for identifying and classifying individual organisms. However, aerial surveys still occur infrequently, and there are often long delays between the acquisition of airborne imagery and its conversion into population monitoring data. Near real‐time monitoring is increasingly important for active management decisions and ecological forecasting. Accomplishing this over large scales requires a combination of airborne imagery, computer vision models to process imagery into information on individual organisms, and automated workflows to ensure that imagery is quickly processed into data following acquisition. Here we present our end‐to‐end workflow for conducting near real‐time monitoring of wading birds in the Everglades, Florida, USA. Imagery is acquired as frequently as weekly using uncrewed aircraft systems (aka drones), processed into orthomosaics (using Agisoft metashape), converted into individual‐level species data using a Retinanet‐50 object detector, post‐processed, archived, and presented on a web‐based visualization platform (using Shiny). The main components of the workflow are automated using Snakemake. The underlying computer vision model provides accurate object detection, species classification, and both total and species‐level counts for five out of six target species (White Ibis, Great Egret, Great Blue Heron, Wood Stork, and Roseate Spoonbill). The model performed poorly for Snowy Egrets due to the small number of labels and difficulty distinguishing them from White Ibis (the most abundant species). By automating the post‐survey processing, data on the populations of these species is available in near real‐time (<1 week from the date of the survey) providing information at the time scales needed for ecological forecasting and active management.

Artificial intelligence correctly classifies developmental stages of monarch caterpillars enabling better conservation through the use of community science photographs

Rapid technological advances and growing participation from amateur naturalists have made countless images of insects in their natural habitats available on global web portals. Despite advances in automated species identification, traits like developmental stage or health remain underexplored or manually annotated, with limited focus on automating these features. As a proof-of-concept, we developed a computer vision model utilizing the YOLOv5 algorithm to accurately detect monarch butterfly caterpillars in photographs and classify them into their five developmental stages (instars). The training data were obtained from the iNaturalist portal, and the photographs were first classified and annotated by experts to allow supervised training of models. Our best trained model demonstrates excellent performance on object detection, achieving a mean average precision score of 95% across all five instars. In terms of classification, the YOLOv5l version yielded the best performance, reaching 87% instar classification accuracy for all classes in the test set. Our approach and model show promise in developing detection and classification models for developmental stages for insects, a resource that can be used for large-scale mechanistic studies. These photos hold valuable untapped information, and we’ve released our annotated collection as an open dataset to support replication and expansion of our methods.

Boosting Point Set-Based Network with Optimal Transport Optimization for Oriented Object Detection

When handling complex remote sensing scenarios, rotational angle information can improve detection accuracy and enhance algorithm robustness, providing support for fine-grained detection. Point set representation is one of the most commonly used methods in arbitrary-oriented object detection tasks, leveraging discrete feature points to represent oriented targets and achieve high accuracy in angle prediction. However, due to the inherent discreteness of point set representation, it is prone to significant impact from isolated points and representational ambiguity in harsh application scenarios, leading to inaccurate detection. To address this issue, an efficient aerial object detector named BE-Det is proposed, which uses the optimal transport (OT) strategy to constrain the positions of isolated points. Additionally, a candidate point set quality evaluation scheme is designed to effectively assess the quality of candidate point sets. Experimental results on two challenging aerial datasets demonstrate that the proposed method outperforms several advanced detection methods.

Technological prototype with artificial intelligence for military security in river environments

Maritime and river security is one of the main concerns of military forces due to the large number of illicit activities that occur. Not to mention the extensive areas that must be monitored, and the weather conditions that can occur. Currently, technologies have become fundamental to leave aside manual surveillance for intelligent systems that allow remote sensing, traffic control, and object detection. Based on the aforementioned problems, the purpose of this research was to design a technological prototype with artificial vision based on an artificial intelligence model to detect water vessels and people in river environments as a support tool for military security. The prototype used at hardware level a Raspberry Pi 3 card and four pre-trained models based on R-CNN, YOLO, EfficientDet and SSD (single shot multibox). The best-performing model was the Mobilnet V2 SSD, with an mAP of 0.83 and an FPS of 5. Finally, this tool can contribute to strengthening the strategic, tactical, and operational capabilities of actors in the military intelligence sector, aimed at protecting sovereignty and territorial integrity to establish an environment of security in society.

Detection of hidden drawings using multi-wavelength dynamic speckle, tuneable algorithms, and unsupervised learning

Abstract We implemented an experiment to reveal hidden drawings on papyrus, utilizing an optical technique based on the speckle phenomenon. The goal is to optimize the detection of hidden objects. Our approach proposes using multiple wavelengths for illumination and tuneable algorithms to process the dynamic speckle images. By implementing the suggested method, we generated various results with varying quality, contingent upon the tuneable algorithm parameters. It's feasible to identify the optimal parameter combination to achieve the most effective visualization of the recovered image. To streamline the selection of tuneable algorithms and mitigate reliance on subjective visual judgment, we employed unsupervised machine learning techniques to determine the conditions necessary to achieve optimal results. This approach simplifies the selection procedure and offers an objective and non-invasive method. Furthermore, the proposed procedure holds promise for extending its application to uncover hidden paintings, subsurface archaeological artefacts, and other dynamic speckle experiments.

Automated acute skin toxicity scoring in a mouse model through deep learning.

This study presents a novel approach to skin toxicity assessment in preclinical radiotherapy trials through an advanced imaging setup and deep learning. Skin reactions, commonly associated with undesirable side effects in radiotherapy, were meticulously evaluated in 160 mice across four studies. A comprehensive dataset containing 7542 images was derived from proton/electron trials with matched manual scoring of the acute toxicity on the right hind leg, which was the target area irradiated in the trials. This dataset was the foundation for the subsequent model training. The two-step deep learning framework incorporated an object detection model for hind leg detection and a classification model for toxicity classification. An observer study involving five experts and the deep learning model, was conducted to analyze the retrospective capabilities and inter-observer variations. The results revealed that the hind leg object detection model exhibited a robust performance, achieving an accuracy of almost 99%. Subsequently, the classification model demonstrated an overall accuracy of about 85%, revealing nuanced challenges in specific toxicity grades. The observer study highlighted high inter-observer agreement and showcased the model's superiority in accuracy and misclassification distance. In conclusion, this study signifies an advancement in objective and reproducible skin toxicity assessment. The imaging and deep learning system not only allows for retrospective toxicity scoring, but also presents a potential for minimizing inter-observer variation and evaluation times, addressing critical gaps in manual scoring methodologies. Future recommendations include refining the system through an expanded training dataset, paving the way for its deployment in preclinical research and radiotherapy trials.

Transformer Based Ship Detector: An Improvement on Feature Map and Tiny Training Set

The exponential increment of commodity exchange has raised the need for maritime border security in recent years. One of the most critical tasks for naval border security is ship detection inside and outside the territorial sea. Conventionally, the task requires a substantial human workload. Fortunately, with the rapid growth of the digital camera and deep-learning technique, computer programs can handle object detection tasks well enough to replace human labor. Therefore, this paper studies how to apply recent state-of-the-art deep-learning networks to the ship detection task. We found that with a suitable number of object queries, the Deformable-DETR method will improve the performance compared to the state-of-the-art ship detector. Moreover, comprehensive experiments on different scale datasets prove that the technique can significantly improve the results when the training sample is limited. Last but not least, feature maps given by the method will focus well on key objects in the image.

Evoattack: suppressive adversarial attacks against object detection models using evolutionary search

Evoattack: suppressive adversarial attacks against object detection models using evolutionary search

Enhanced lightweight infrared object detection algorithm for assistive navigation in visually impaired individuals

Enhanced lightweight infrared object detection algorithm for assistive navigation in visually impaired individuals

Inhomogeneous Illumination Image Enhancement Under Extremely Low Visibility Condition

Imaging through dense fog presents unique challenges, with essential visual information crucial for applications like object detection and recognition, thereby hindering conventional image processing methods. Despite improvements through neural network-based approaches, these techniques falter under extremely low visibility conditions exacerbated by inhomogeneous illumination, which degrades deep learning performance due to inconsistent signal intensities. We introduce in this paper a novel method that adaptively filters background illumination based on Structural Differential and Integral Filtering (SDIF) to enhance only the vital signal information. The grayscale banding is eliminated by incorporating a visual optimization strategy based on image gradients. Maximum Histogram Equalization (MHE) is used to achieve high contrast while maintaining fidelity to the original content. We evaluated our algorithm using data collected from both a fog chamber and outdoor environments and performed comparative analyses with existing methods. Our findings demonstrate that our proposed method significantly enhances signal clarity under extremely low visibility conditions and out-performs existing techniques, offering substantial improvements for deep fog imaging applications.

Underwater Object Detection Algorithm Based on an Improved YOLOv8

Due to the complexity and diversity of underwater environments, traditional object detection algorithms face challenges in maintaining robustness and detection accuracy when applied underwater. This paper proposes an underwater object detection algorithm based on an improved YOLOv8 model. First, the introduction of CIB building blocks into the backbone network, along with the optimization of the C2f structure and the incorporation of large-kernel depthwise convolutions, effectively enhances the model’s receptive field. This improvement increases the capability of detecting multi-scale objects in complex underwater environments without adding a computational burden. Next, the incorporation of a Partial Self-Attention (PSA) module at the end of the backbone network enhances model efficiency and optimizes the utilization of computational resources while maintaining high performance. Finally, the integration of the Neck component from the Gold-YOLO model improves the neck structure of the YOLOv8 model, facilitating the fusion and distribution of information across different levels, thereby achieving more efficient information integration and interaction. Experimental results show that YOLOv8-CPG significantly outperforms the traditional YOLOv8 in underwater environments. Precision and Recall show improvements of 2.76% and 2.06%. Additionally, mAP50 and mAP50-95 metrics have increased by 1.05% and 3.55%, respectively. Our approach provides an efficient solution to the difficulties encountered in underwater object detection.

Design of apple recognition model based on improved deep learning object detection framework Faster-RCNN

It the stage of digital economy development in China, intelligent recognition technology is used in agriculture, forestry and planting industries. This paper improves and optimizes apple recognition based on Faster-RCNN, a deep learning target detection framework, and analyzes the advantages and disadvantages of the improved target detection and recognition model. This study compares and analyzes the conventional Faster-RCNN apple recognition model and the enhanced Faster-RCNN apple recognition model design based on the deep learning target detection framework. From the two groups of recognition model systems under different design schemes, the research and analysis of comprehensive performance, sensitivity and coupling shows that the improved Faster-RCNN Apple recognition model has higher accuracy and more accurate positioning for Apple recognition. The results show that the improved Faster-RCNN apple recognition model can more effectively improve the design of the apple recognition model at the present stage, reduce labor costs, and significantly improve work efficiency. It also promotes the development of the domestic digital economy and provides research support and value for future intelligence and deep learning development paths.

YOLOX-S-TKECB: A Holstein Cow Identification Detection Algorithm

Accurate identification of individual cow identity is a prerequisite for the construction of digital farms and serves as the basis for optimized feeding, disease prevention and control, breed improvement, and product quality traceability. Currently, cow identification faces challenges such as poor recognition accuracy, large data volumes, weak model generalization ability, and low recognition speed. Therefore, this paper proposes a cow identification method based on YOLOX-S-TKECB. (1) Based on the characteristics of Holstein cows and their breeding practices, we constructed a real-time acquisition and preprocessing platform for two-dimensional Holstein cow images and built a cow identification model based on YOLOX-S-TKECB. (2) Transfer learning was introduced to improve the convergence speed and generalization ability of the cow identification model. (3) The CBAM attention mechanism module was added to enhance the model’s ability to extract features from cow torso patterns. (4) The alignment between the apriori frame and the target size was improved by optimizing the clustering algorithm and the multi-scale feature fusion method, thereby enhancing the performance of object detection at different scales. The experimental results demonstrate that, compared to the traditional YOLOX-S model, the improved model exhibits a 15.31% increase in mean average precision (mAP) and a 32-frame boost in frames per second (FPS). This validates the feasibility and effectiveness of the proposed YOLOX-S-TKECB-based cow identification algorithm, providing valuable technical support for the application of dairy cow identification in farms.

YOLOv5DA: An Improved YOLOv5 Model for Posture Detection of Grouped Pigs

Accurate posture detection is the foundation for analyzing animal behavior, which can promote animal welfare. With the development of computer vision, such technology has been widely used in analyzing animal behavior without physical contact. However, computer vision technology for pig posture detection often suffers from problems of missed or false detection due to complex scenarios. To solve the problem, this study proposed a novel object detection model YOLOv5DA, which was based on YOLOv5s and designed for pig posture detection from 2D camera video. Firstly, we established the annotated dataset (7220 images) including the training set (5776 images), validation set (722 images), and test set (722 images). Secondly, an object detection model YOLOv5DA based on YOLOv5s was proposed to recognize pig postures (standing, prone lying, and side lying), which incorporated Mosaic9 data augmentation, deformable convolution, and adaptive spatial feature fusion. The comparative and ablation experiments were conducted to verify the model’s effectiveness and reliability. Finally, we used YOLOv5DA to detect the posture distribution of pigs. The results revealed that the standing posture was more frequent in the morning and afternoon and the side-lying posture was most common at noon. This observation demonstrated that the posture of pigs is influenced by temperature variations. The study demonstrated that YOLOv5DA could accurately identify three postures of standing, prone lying, and side lying with an average precision (AP) of 99.4%, 99.1%, and 99.1%, respectively. Compared with YOLOv5s, YOLOv5DA could effectively handle occlusion while increasing the mean precision (mAP) by 1.7%. Overall, our work provided a highly accurate, effective, low-cost, and non-contact strategy of posture detection in grouped pigs, which can be used to monitor pig behavior and assist in the early prevention of disease.