OATCR: Outdoor Autonomous Trash-Collecting Robot Design Using YOLOv4-Tiny

<title>Abstract</title> The future of agriculture is intertwined with automation. Accurate fruit detection, yield estimation, and harvest time prediction are crucial for efficient supply chain management by optimizing resources and logistic utilization. Computer vision can automate these tasks to reduce labour costs and improve efficiency by training deep learning models on appropriate data to perform knowledge-based tasks. Although fruit detection has been the focus of literature, yield prediction and harvest time estimation remain practical challenges for farmers. This is particularly important for high-value, highly perishable crops, such as raspberries, where contractual obligations require precise harvest timing. This paper addresses this gap by providing a learning-based framework for raspberry maturity detection and harvest time estimation. For this end, a phenology study of raspberry plants from three different cultivars was carried out and seven development stages were identified based on the BBCH-scale; for the first time, a phenology-based dataset of developing raspberry flowers and fruit was curated and made available publicly, which contains 1,853 high-resolution images and 6,907 manually labelled annotations. A comprehensive benchmark was developed from state-of-the-art object detection models, and finally, a tailored deep learning model capable of real-time inference was established that achieved 92.2% detection accuracy in the vertical farm field test. This paper provides a model that can estimate raspberry yield 28 to 33 days in advance of harvesting, a dataset that can be used by other researchers fpr object detection and segmentation model development, and a framework that can be extended for precise yield and harvest time estimation.

Sea turtles exhibit high migratory rates and occupy a broad range of habitats, which in turn makes monitoring these taxa challenging. Applying deep learning (DL) models to vast image datasets collected from citizen science programs can offer promising solutions to overcome the challenge of monitoring the wide habitats of wildlife, particularly sea turtles. Among DL models, object detection models, such as the You Only Look Once (YOLO) series, have been extensively employed for wildlife classification. Despite their successful application in this domain, detecting objects in images with complex backgrounds, including underwater environments, remains a significant challenge. Recently, instance segmentation models have been developed to address this issue by providing more accurate classification of complex images compared to traditional object detection models. This study compared the performance of two state-of-the-art DL methods namely; the object detection model (YOLOv5) and instance segmentation model (YOLOv5-seg), to detect and classify sea turtles. The images were collected from iNaturalist and Google and then divided into 64% for training, 16% for validation, and 20% for test sets. Model performance during and after finishing training was evaluated by loss functions and various indexes, respectively. Based on loss functions, YOLOv5-seg demonstrated a lower error rate in detecting rather than classifying sea turtles than the YOLOv5. According to mean Average Precision (mAP) values, which reflect precision and recall, the YOLOv5-seg model showed superior performance than YOLOv5. The mAP0.5 and mAP0.5:0.95 for the YOLOv5 model were 0.885 and 0.795, respectively, whereas for the YOLOv5-seg, these values were 0.918 and 0.831, respectively. In particular, based on the loss functions and classification results, the YOLOv5-seg showed improved performance for detecting rather than classifying sea turtles compared to the YOLOv5. The results of this study may help improve sea turtle monitoring in the future.

In the "blind review" mode for technical bid evaluation within the power industry, review experts must make impartial judgments based on anonymized technical response documents. However, these documents often contain a large amount of complex graphical content, which existing general Optical Character Recognition (OCR) technology struggles to accurately identify. To address this, this study first constructs a specialized image dataset covering typical power industry response document scenarios. Subsequently, an improved You Only Look Once (YOLO) object detection model is developed, incorporating an anchor box clustering strategy optimized for small, densely packed power symbols and a multi-scale feature fusion mechanism. Then, a local contextual relationship inference module based on Graph Neural Networks is designed and integrated. Finally, an end-to-end recognition pipeline from image input to structured information output is constructed. Experimental results indicate that the improved model achieves a mean Average Precision (mAP) of 90.0% in key equipment detection tasks; in topology relationship restoration experiments, the average Graph Edit Distance (GED) is 0.172; in end-to-end efficiency experiments simulating real blind review scenarios, the average processing time per document is only 8.7 seconds, with an information completeness rate of 98.2%. The research demonstrates that the YOLO-based object detection and relationship modeling method, customized for the blind review mode, can effectively address the challenge of recognizing complex image content in power response documents, providing reliable technical support for the efficiency and fairness of blind reviews.

In this paper, we investigate the problem of placing sensors in a mobile social network to get quickly informed about contagious outbreaks, i.e., placing k sensors in a network in order to minimize the time until a contaminant - starting from a random node in the network - is detected. We aim to optimize the Sensor Placement from two complementary directions. One is to improve the original greedy algorithm and its extensions [13] to reduce sensor selection time, and the other is to propose a new Quickest Path heuristic that can shorten the detection time. We test and compare our algorithms with previous algorithms on four real data sets. Experimental results show that 1) the new greedy algorithm is more efficient than existing greedy algorithms in terms of selection time, 2) the quickest path heuristic obtains less detection time than centrality-based heuristics, and is as effective as the greedy algorithms, and 3) the new heuristic has the potential to scale well to large networks, having low detection time and selection time.

An adversarial attack on DNN-based black-box object detectors

Recently, edge computing has emerged as a prevailing paradigm in applying deep learning-based object detection models, offering a promising solution for time-sensitive tasks. However, existing edge object detection faces several challenges: 1) These methods struggle to balance detection precision and model lightweightness. 2) Existing generalized edge-deployment designs offer limited adaptability for object detection. 3) Current works lack real-world evaluation and validation. To address these challenges, we propose the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><u>E</u>dge <u>D</u>etection <u>Toolbox</u></i> (ED-TOOLBOX), which leverages generalizable plug-and-play components to enable edge-site adaptation of object detection models. Specifically, we propose a lightweight <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Reparameterized Dynamic Convolutional Network</i> (Rep-DConvNet) that employs a weighted multi-shape convolutional branch structure to enhance detection performance. Furthermore, ED-TOOLBOX includes a <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Sparse Cross-Attention</i> (SC-A) network that adopts a localized-mapping-assisted self-attention mechanism to facilitate a well-crafted <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Joint Module</i> in adaptively transferring features for further performance improvement. Moreover, we propose an <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Efficient Head</i> for the classification and location modules to achieve more efficient prediction. Additionally, in practical industrial scenarios, we identify that helmet detection-one of the most representative edge object detection tasks-overlooks band fastening, which introduces potential safety hazards. To address this, we build a <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Helmet Band Detection Dataset</i> (HBDD) and apply an edge object detection model optimized by the ED-TOOLBOX to tackle this real-world task. Extensive experiments validate the effectiveness of components in ED-TOOLBOX. In visual surveillance simulations, ED-TOOLBOX-assisted edge detection models outperform six <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">state-of-the-art</i> methods, enabling real-time and accurate detection. These results demonstrate that our approach offers a superior solution for edge object detection.

Supramolecular chemistry relies on the dynamic association/dissociation of molecules through non-covalent interactions. These interactions of a self-assembled system can be strategically exploited for sensing several microorganisms. Moreover, supramolecular systems can also be combined with other functional components like nanoparticles, self-assembled monolayers, and microarray systems to produce multicomponent sensors with higher sensitivity and lower detection time. In this review, we will discuss how cutting-edge supramolecular chemistry has enabled scientists to develop microbial biosensors with high reliability and rapid detection time. Moreover, they produce high-throughput operations, real-time monitoring, extensive operation platforms, and cost-effective production. This review can serve as a conceptual background for understanding state-of-the-art rapid detection methods of microbial biosensing.

Objective To explore the detection of cortical responses to continuous speech using a single EEG channel. Particularly, to compare detection rates and times using a cross-correlation approach and parameters extracted from the temporal response function (TRF). Design EEG from 32-channels were recorded whilst presenting 25-min continuous English speech. Detection parameters were cross-correlation between speech and EEG (XCOR), peak value and power of the TRF filter (TRF-peak and TRF-power), and correlation between predicted TRF and true EEG (TRF-COR). A bootstrap analysis was used to determine response statistical significance. Different electrode configurations were compared: Using single channels Cz or Fz, or selecting channels with the highest correlation value. Study sample Seventeen native English-speaking subjects with mild-to-moderate hearing loss. Results Significant cortical responses were detected from all subjects at Fz channel with XCOR and TRF-COR. Lower detection time was seen for XCOR (mean = 4.8 min) over TRF parameters (best TRF-COR, mean = 6.4 min), with significant time differences from XCOR to TRF-peak and TRF-power. Analysing multiple EEG channels and testing channels with the highest correlation between envelope and EEG reduced detection sensitivity compared to Fz alone. Conclusions Cortical responses to continuous speech can be detected from a single channel with recording times that may be suitable for clinical application.

Foreign bodies such as a surgical gauze can be retained in the body after surgery and in some cases cannot be detected by postoperative radiography. The aim of this study was to develop an object detection model capable of postsurgical detection of retained gauze in the body. The object detection model used deep learning using abdominal radiographs, and a phantom study was performed to evaluate the ability of the model to automatically detect retained surgical gauze. The object detection model was constructed using a Single Shot MultiBox Detector (SSD) 300. In total, 268 abdominal phantom images were used: 180 gauze images were used as training data, 20 gauze images were used as validation data, and an additional 34 gauze images and 34 nongauze images were used as test data. To evaluate the performance of the object detection model, a confusion matrix was created and the accuracy and sensitivity were calculated. True-positive (TP) rate, true-negative (TN) rate, false-positive (FP) rate, and false-negative (FN) rate were 0.92, 1.00, 0.00, and 0.08, respectively. Accuracy was 0.96, and sensitivity was 0.92. The object detection model could detect surgical gauze on abdominal phantom images with a high accuracy and sensitivity.

I. ABSTRACT Fire hazards pose a significant threat to life and property, making early detection crucial for minimizing damage. This project focuses on AI-based fire detection and alarming using computer vision and deep learning techniques. The system employs CCTV cameras or laptop cameras to capture real-time video feeds, which are analysed using Python, OpenCV, and deep learning models to detect fire accurately. A Convolutional Neural Network (CNN) or YOLO (You Only Look Once) object detection model is trained on fire datasets to distinguish fire from non-fire scenarios, reducing false alarms. Once fire is detected, the system triggers an alarm sound and can also send notifications via SMS, email, or IoT-based alerts to concerned authorities. This AI-powered solution is designed for high accuracy and efficiency, making it suitable for smart surveillance, industrial safety, and residential fire monitoring. The integration of edge computing devices like Raspberry Pi or Jetson Nano can further enhance real-time processing. The proposed system aims to improve fire safety measures by providing automated, real-time fire detection, reducing response time, and preventing large-scale disasters. II. KEYWORDS Fire Detection, Real-time Fire Monitoring, Computer Vision (CV).

The food industry increasingly depends on technological assets to improve the efficiency and accuracy of fruit processing and quality control. This article enhances the application of computer vision with collaborative robotics to create a non-destructive system. The system can automate the detection and handling of fruits, particularly tomatoes, reducing the reliance on manual labor and minimizing damage during processing. This system was developed with a Raspberry Pi 5 to capture images of the fruit using a PiCamera module 3. After detecting the object, a command is sent to a Universal Robotics UR3e robotic arm via Ethernet cable, using Python code that integrates company functions and functions developed specifically for this application. Four object detection models were developed using the TensorFlow Object Detection API, converted to TensorFlow Lite, to detect two types of fruit (tomatoes) using deep learning techniques. Each fruit had two versions of the models. The models obtained 67.54% mAP for four classes and 64.66% mAP for two classes, A rectangular work area was created for the robotic arm and computer vision to work together. After 640 manipulation tests, a reliable area of 262 × 250 mm was determined for operating the system. In fruit sorting facilities, this system can be employed to automatically classify fruits based on size, ripeness, and quality. This ensures consistent product standards and reduces waste by sorting fruits according to pre-defined criteria. The system’s ability to detect multiple fruit types with high accuracy enables it to integrate into existing workflows, thereby increasing productivity and profitability for food processing companies. Additionally, the non-destructive nature of this technology allows for the inspection of fruits without causing any damage, ensuring that only the highest-quality produce is selected for further processing. This application can enhance the speed and precision of quality control processes, leading to improved product quality and customer satisfaction.

Backdoor Attacks with Wavelet Embedding: Revealing and enhancing the insights of vulnerabilities in visual object detection models on transformers within digital twin systems

Computer vision techniques are becoming increasingly popular for monitoring pig behavior. For instance, object detection models allow us to detect the presence of pigs, their location, and their posture. The performance of object detection models can be affected by variations in lighting conditions (e.g., intensity, spectrum, and uniformity). Furthermore, lighting conditions can influence pigs' active and resting behavior. In the context of experiments testing different lighting conditions, a detection model was developed to detect the location and postures of group-housed growing-finishing pigs. The objective of this paper is to validate the model developed using YOLOv8 detecting standing, sitting, sternal lying, and lateral lying pigs. Training, validation, and test datasets included annotation of pigs from 10 to 24wk of age in 10 different light settings; varying in intensity, spectrum, and uniformity. Pig detection was comparable for the different lighting conditions, despite a slightly lower posture agreement for warm light and uneven light distribution, likely due to a less clear contrast between pigs and their background and the presence of shadows. The detection reached a mean average precision (mAP) of 89.4%. Standing was the best-detected posture with the highest precision, sensitivity, and F1 score, while the sensitivity and F1 score of sitting was the lowest. This lower performance resulted from confusion of sitting with sternal lying and standing, as a consequence of the top camera view and a low occurrence of sitting pigs in the annotated dataset. This issue is inherent to pig behavior and could be tackled using data augmentation. Some confusion was reported between types of lying due to occlusion by pen mates or pigs' own bodies, and grouping both types of lying postures resulted in an improvement in the detection (mAP = 97.0%). Therefore, comparing resting postures (both lying types) to active postures could lead to a more reliable interpretation of pigs' behavior. Some detection errors were observed, e.g., two detections for the same pig were generated due to posture uncertainty, dirt on cameras detected as a pig, and undetected pigs due to occlusion. The localization accuracy measured by the intersection over union was higher than 95.5% for 75% of the dataset, meaning that the location of predicted pigs was very close to annotated pigs. Tracking individual pigs revealed challenges with ID changes and switches between pen mates, requiring further work.

In this project, we develop and apply a Smart Water Dispenser system, which is combined with object recognition and fluid level control supported by Ultrasonic Sensors, Raspberry Pi, and also DC Motors. The essence of this system is to develop a system using the Raspberry Pi 4 Model B with other components that have been integrated and interrelated Hardware and programming using OpenCV, YOLO V8, and other components, the point is that the cup can be detected, and water filling is done precisely and automatically. The process carried out is the detection of cups automatically using Raspberry Pi which is in charge of controlling the DC Motor and also the Ultrasonic sensor (HC-SR04) and detecting based on the volume of water with precision. The dispenser functions to pump water based on the condition of the volume of water in the glass and stop pumping if the volume of the glass has been fulfilled, aka not spilling with a percentage of 90%. In the scenario process, the cup search process is first carried out by scanning three times until a cup is found, if a cup is found, then the sensor component and the valve for the release of water in the hose will stop right at the position of the cup and the water will fill the cup automatically. Otherwise, the system will move backward and the system will be turned off. The first testing process has been successful and shows the effectiveness of the system in the process of finding cups and managing water levels. This innovation shows hope for improving user comfort, especially for disabilities, utilizing advanced technology in object recognition, and of course, saving water usage. In the testing process obtained 95% to 97% accuracy in object detection with different types of cups.

The global expansion of visual impairment is a very sensitive subject. Living in constant uncertainty has always had a negative impact on many aspects of their lives. The digital revolution is accelerating in the fast-paced culture, and it has had a good impact on blind people's life as well. This study intends to develop a support system for those who are blind or visually impaired. People with visual impairments from the partially unsighted to the completely blind can use the image recognition and text navigation system to support and direct them via auditory commands. The tool includes the sequential operations of indoor obstacle detection and recognition. The presence of an object is identified using an ultrasonic sensor. Here, a single chip known as a Raspberry Pi is being integrated with both the text navigation device and the identification system. The Raspberry Pi operates with great accuracy and speed. An object is photographed by the navigation system's camera module, processed using the single shot detection (SSD) algorithm, and then converted to text by an optical character recognition (OCR) engine before being communicated over the speaker using the text-to-speech (TTS) converter tool. The object recognition and the text navigation system are specially designed for blind individuals to operate the tool conveniently in a solitary manner.