Affect Detection Performance Research Articles

Unharvested Palm Fruit Bunch Ripeness Detection with Hybrid Color Correction

Unharvested palm fruit bunch (PFB) ripeness detection is crucial for efficient palm oil production as labor costs rise and worker shortages grow. Challenges remain under varying illumination, affecting detection performance. Although the deep learning-based color constancy model shows promise, the limitation lies in the lack of data. This paper addresses these issues by proposing a hybrid color correction method that combines the learning-based and the physical-based color constancy technique, then trains with YOLOv8 to enhance unharvested PFB detection performance. Specifically, the captured images and their corresponding simultaneous spectra are used to generate the ground truth images, then map the images to other ambient spectra to train the color constancy model, and finally, train the YOLOv8 with the corrected images for ripeness detection. Results demonstrate a 1.5% improvement in the mean Average Precision (mAP) @0.5 of YOLOv8, increasing from 88.2% to 89.7% after applying the proposed hybrid color correction method. For each ripeness category, the mAP@0.5 for unripe and ripe bunches increased by 1.7% and 1.8%, respectively. Underripe bunches exhibit the most increment in mAP of 2.1%, while the overripe bunch gives the least increase of 0.4%. The qualitative results show that the model has enhanced the ability to distinguish subtle differences in PFB ripeness. Consequently, it advances the effectiveness and reliability of PFB ripeness detection in practical palm oil production scenarios, especially for unripe to ripe bunches wherein color gives the most information. This research demonstrates potential in object detection tasks that incorporate illumination-based color correction methods.

Miniaturized optical system of a laser seeker based on a dynamic and static performance co-design

The laser semi-active seeker, crucial for miniaturizing guided munitions, offers high precision, anti-jamming, and rapid response. Current research focuses on system response, spot parameters, and angular velocity accuracy. With small-caliber munitions on the rise, further miniaturization and weight reduction are essential. This paper proposes a design for an optical system that balances dynamic imaging quality and miniaturized static structure. Using quadrant detector principles, factors affecting detection performance are analyzed to determine the optimal spot size and imaging range. Based on imaging requirements and the optical properties of a dual-reflector structure, we determine the design parameters of the system to achieve a combination of dynamic and static performance. Evaluations using ZEMAX and MATLAB confirm that the system achieves ±4∘ linear and ±8∘ full fields of view, 0.12% distortion, 4.52 mrad angular accuracy, a 15.72 mm axial length, and has the possibility of actual production. The system is compact, with excellent spot roundness, stability, and uniform energy distribution, providing a reference for future guided weapon research.

FDADNet: Detection of Surface Defects in Wood-Based Panels Based on Frequency Domain Transformation and Adaptive Dynamic Downsampling

The detection of surface defects on wood-based panels plays a crucial role in product quality control. However, due to the complex background and low contrast of defects in wood-based panel images, features extracted by traditional deep learning methods based on spatial domain processing often contain noise and blurred boundaries, which severely affects detection performance. To address these issues, we have proposed a wood-based panel surface defect detection method based on frequency domain transformation and adaptive dynamic downsampling (FDADNet). Specifically, we designed a Multi-axis Frequency Domain Weighted Information Representation Module (MFDW), which effectively decoupled the indistinguishable low-contrast defects from the background in the transform domain. Gaussian filtering was then employed to eliminate noise and blur between the defects and the background. Additionally, to tackle the issue of scale differences in defects that led to difficulties in accurate capture, we designed an Adaptive Dynamic Convolution (ADConv) module for downsampling. This method flexibly compressed and enhanced features, effectively improving the differentiation of the features of objects of varying scales in the transform space, and ultimately achieved effective defect detection. To compensate for the lack of data, we constructed a dataset of wood-based panel surface defects, WBP-DET. The experimental results showed that the proposed FDADNet effectively improved the detection performance of wood-based panel surface defects in complex scenarios, achieving a solid balance between efficiency and accuracy.

DESIGN OF IMAGE PROCESSING APPLICATION FOR FOREIGN OBJECT INSPECTION USING DRONE ON AIRCRAFT RUNWAYS

Flight safety is a crucial aspect of the aviation industry, heavily reliant on optimal runway conditions for successful takeoffs and landings. Foreign Object Debris (FOD) such as stones and potholes on the runway pose significant threats, necessitating preventive measures to avoid operational failures and aircraft damage. This research focuses on designing an image processing application for drone usage to enhance runway inspections. The Research and Development (R&D) method is employed from application design to field testing. Image processing utilizes the You Only Look Once (YOLO) method for swift and accurate FOD detection, maximizing drone effectiveness in identifying risks. Findings indicate that object size and image capture distance significantly affect detection performance. This analysis provides a basis for model optimization through improved data augmentation techniques and parameter adjustments to address challenges in detecting objects of various sizes.

Detection range of a passing submerged emitter of sound in ambient noisea).

An underwater acoustic detection problem is studied in which the ambient noise present at a receiver is calculated, given information describing environmental conditions, including windspeed, and the positions of nearby ships which act as sources of background noise. The signal, whose detection is sought, is narrowband and transmitted from a source that passes by the receiver along a straight track. Cumulative Probability of Detection (CPoD) is calculated along a series of tracks with increasing closest-point-of-approach distances to the receiver. Two detection ranges are analyzed, a so-called "defender" detection range and an "intruder" detection range. Both are conservative measures associated with CPoD equaling 0.5 during the transit of the submerged vessel. Predictions of the detection range are compared across independent attempts to solve the same problem with different modelling approaches. The spread of results (i.e., the "reproducibility" of the predictions) is discussed and reasons for differences are highlighted. Environmental conditions that strongly affect detection performance are discussed, as is the use of CPoD as a single-valued metric to describe detection performance.

Enhancing Hyperspectral Anomaly Detection with a Novel Differential Network Approach for Precision and Robust Background Suppression

The existing deep-learning-based hyperspectral anomaly detection methods detect anomalies by reconstructing a clean background. However, these methods model the background of the hyperspectral image (HSI) through global features, neglecting local features. In complex background scenarios, these methods struggle to obtain accurate background priors for training constraints, thereby limiting the anomaly detection performance. To enhance the capability of the network in extracting local features and improve anomaly detection performance, a hyperspectral anomaly detection method based on differential network is proposed. First, we posit that anomalous pixels are challenging to be reconstructed through the features of surrounding pixels. A differential convolution method is introduced to extract local punctured neighborhood features in the HSI. The differential convolution contains two types of kernels with different receptive fields. These kernels are adopted to obtain the outer window features and inner window features. Second, to improve the feature extraction capability of the network, a local detail attention and a local Transformer attention are proposed. These attention modules enhance the inner window features. Third, the obtained inner window features are subtracted from the outer window features to derive differential features, which encapsulate local punctured neighborhood characteristics. The obtained differential features are employed to reconstruct the background of the HSI. Finally, the anomaly detection results are extracted from the difference between the input HSI and the reconstructed background of the HSI. In the proposed method, for each receptive field kernel, the optimization objective is to reconstruct the input HSI rather than the background HSI. This way circumvents problems where the background constraint biases might affect detection performance. The proposed method offers researchers a new and effective approach for applying deep learning in a local area to the field of hyperspectral anomaly detection. The experiments are conducted with multiple metrics on five real-world datasets. The proposed method outperforms eight state-of-the-art methods in both subjective and objective evaluations.

Low-Frequency Blast Detection Using a Large-N Dark Fiber in Noisy Environments: Template Matching and Optimal Channel Selection

Abstract Distributed acoustic sensing (DAS), deployed on dark telecom fiber, is well-positioned to play a significant role in seismic monitoring networks because of the combination of a large aperture, fine spatial resolution, broadband sensitivity, and the ubiquitous presence of unused telecommunication fibers in many areas of the world. In this study, we explore the feasibility of dark-fiber array deployed in a noisy environment for detecting small explosions. We test the effectiveness of template matching for the detection of low-frequency blasts generated by mining activities in the Imperial Valley, California. We first evaluate dark-fiber detection performance by analyzing the relationship between detection threshold (DT) and the number of DAS channels used. We find that although, as expected, increasing the number of channels yields higher detection significance and lowers DT, the gain in performance is far from linear, with local anomalies across the DAS cable associated with zones of higher noise. We focus on investigating the types of noise affecting template matching and practical approaches mitigating anthropogenic noise that lower detection performance. Using median absolute deviation, we identify two types of noise sources affecting detection performance. Next, we design a voting scheme that selects DAS channels contributing to lowering of the DT and ensures improvement in detection when adding sequential channels. Finally, we compare dark-fiber detection performance with nearby conventional seismometers and find that a single station can outperform up to ∼10 DAS channels. However, using the full aperture of our dark-fiber transect allows to obtain ∼10% lower DT and yields fewer false-positive detections than an array of four seismometers. Methodological solutions for noise assessment and channel selection allow us to fully benefit from the large aperture and dense sampling offered by dark fiber. The findings of this study are a step toward incorporating existing telecom fibers into novel explosion-monitoring workflows.

Design and analysis of novel high-performance thick epitaxial SiC radiation detector with low operating voltage by multiple p-type buried layers

In this work, epitaxial silicon carbide (SiC) radiation detectors (E-SRD) were modeled, and SiC PiN diodes with multiple p-type buried layers (PBL), named MPBL PiN, were proposed as E-SRD to address the problem that detection performance of E-SRD using convention PiN diodes (Cov. PiN) with thick epilayer is limited by its high fully depleted voltage (VFD). The charge collection efficiency (CCE), pulse height and relevant physical mechanism affecting detection performance of E-SRD were investigated. When both Cov. PiN and MPBL PiN have a 100 μm epilayer with n-type doping concentration of 3 × 1014 cm−3, the VFD of Cov. PiN reaches 2800 V, while that for MPBL PiN is only 550 V. 14 MeV neutron detection simulation results showed that, compared to Cov. PiN, the CCE, pulse height and radiation resistance of MPBL PiN improve significantly under low operating voltage within 1000 V, meanwhile its lower near-surface electric field is conducive to reducing leakage. Further, since the VFD of MPBL PiN is proportional to the thickness of the epilayer (HE) instead of to the square of the HE like Cov. PiN, the VFD of MPBL PiN with ultra-thick epilayer is much lower than that of Cov. PiN. Therefore, under commonly used operating voltage within 1000V, the sensitivity and pulse height of MPBL PiN can be proportional to the HE, while that of Cov. PiN hardly improved when HE is above 100 μm. So MPBL PiN can give full play to the high-performance detection advantages of ultra-thick epilayer, which is expected to promote the development and application of ultra-thick epilayer E-SRD with low operating voltage.

Warning: Humans cannot reliably detect speech deepfakes.

Speech deepfakes are artificial voices generated by machine learning models. Previous literature has highlighted deepfakes as one of the biggest security threats arising from progress in artificial intelligence due to their potential for misuse. However, studies investigating human detection capabilities are limited. We presented genuine and deepfake audio to n = 529 individuals and asked them to identify the deepfakes. We ran our experiments in English and Mandarin to understand if language affects detection performance and decision-making rationale. We found that detection capability is unreliable. Listeners only correctly spotted the deepfakes 73% of the time, and there was no difference in detectability between the two languages. Increasing listener awareness by providing examples of speech deepfakes only improves results slightly. As speech synthesis algorithms improve and become more realistic, we can expect the detection task to become harder. The difficulty of detecting speech deepfakes confirms their potential for misuse and signals that defenses against this threat are needed.

Detection by radar of a slow‐moving target on the sea surface

AbstractA simple model is proposed for the radar returns from a slow‐moving target with a wake on the sea surface, based on measurements of returns from a submarine mast. This model is used to investigate the effect of a wake with a broad Doppler spectrum on the detection of a slow‐moving target in sea clutter. It is shown how the relative magnitudes of the returns from the target and its wake affect detection performance. Non‐coherent and coherent detectors are compared and it is shown that in some circumstance non‐coherent detectors may give better performance, dependent on the choice of radar waveform parameters.

Multitarget Intelligent Recognition of Petrographic Thin Section Images Based on Faster RCNN

The optical features of mineral composition and texture in petrographic thin sections are an important basis for rock identification and rock evolution analysis. However, the efficiency and accuracy of human visual interpretation of petrographic thin section images have depended on the experience of experts for a long time. The application of image-based computer vision and deep-learning algorithms to the intelligent analysis of the optical properties of mineral composition and texture in petrographic thin section images (in plane polarizing light) has the potential to significantly improve the efficiency and accuracy of rock identification and classification. This study completed the transition from simple petrographic thin image classification to multitarget detection, to address more complex research tasks and more refined research scales that contain more abundant information, such as spatial, quantitative and category target information. Oolitic texture is an important paleoenvironmental indicator that widely exists in sedimentary records and is related to shallow water hydraulic conditions. We used transfer learning and image data augmentation in this paper to identify the oolitic texture of petrographic thin section images based on the faster region-based convolutional neural network (Faster RCNN) method. In this study, we evaluated the performance of Faster RCNN, a two-stage object detection algorithm, using VGG16 and ResNet50 as backbones for image feature extraction. Our findings indicate that ResNet50 outperformed VGG16 in this regard. Specifically, the Faster RCNN model with ResNet50 as the backbone achieved an average precision (AP) of 92.25% for the ooids test set, demonstrating the accuracy and reliability of this approach for detecting ooids. The experimental results also showed that the uneven distribution of training sample images and the complexity of images both significantly affect detection performance; however, the uneven distribution of training sample images has a greater impact. Our work is preliminary for intelligent recognition of multiple mineral texture targets in petrographic thin section images. We hope that it will inspire further research in this field.

Hydrogel-Film-Fabricated Fluorescent Biosensors with Aggregation-Induced Emission for Albumin Detection through the Real-Time Modulation of a Vortex Fluidic Device.

Hydrogels have various promising prospects as a successful platform for detecting biomarkers, and human serum albumin (HSA) is an important biomarker in the diagnosis of kidney diseases. However, the difficult-to-control passive diffusion kinetics of hydrogels is a major factor affecting detection performance. This study focuses on using hydrogels embedded with aggregation-induced emission (AIE) fluorescent probe TC426 to detect HSA in real time. The vortex fluidic device (VFD) technology is used as a rotation strategy to control the reaction kinetics and micromixing during measurement. The results show that the introduction of VFD could significantly accelerate its fluorescence response and effectively improve the diffusion coefficient, while VFD processing could regulate passive diffusion into active diffusion, offering a new method for future sensing research.

Impact of angle on photothermal radiometry and modulated luminescence (PTR/LUM) value

ObjectiveTo evaluate the impact of scanning angles to detect/quantify non-cavitated caries by photothermal-radiometry and modulated-luminescence (PTR/LUM, Canary System) and to evaluate the association of PTR/LUM value with lesion depth (LD), including sound tissue thickness under the lesion (ST). MethodsThirty human extracted premolars were selected based on micro-computed tomography [μ-CT: sound (n=12), lesions into outer-half of enamel (n=6), lesions into inner-half of enamel (n=6), lesions into outer one-third of dentine (n=6)]. Each tooth sample was scanned 90° directly contacted to the center of non-cavitated lesion or sound smooth surface, and tilted 10° and 20° in four directions: buccal/lingual/occlusal/cervical. The procedure was repeated 48 h later. Lesion depth and ST [ST=5000 µm (maximum PTR/LUM scanning depth)-LD] were measured at the same scanning direction on μ-CT images. Sensitivity, specificity, area under the Receiver Operating Characteristic curve (AUC), and intraclass correlation coefficients (ICC) for different scanning angles were calculated. Sensitivity was further evaluated based on lesion extensions. Relationships between PTR/LUM value and lesion depth, and between PTR/LUM value and LD/ST-Ratio were evaluated. ResultsPTR/LUM value showed significant differences among scanning angles. Overall sensitivity (78%–89%), specificity (66%–87%), AUC (0.86–0.92) and ICC (0.89–0.99), sensitivity based on lesion extensions presented no significant differences among angles. PTR/LUM value showed moderate correlations (0.56–0.74) with deepest lesion depth and LD/ST-Ratios. ConclusionThe scanning angle within 20° increments might impact PTR/LUM value statistically; however, it did not affect PTR/LUM detection performance. PTR/LUM values were positively correlated with non-cavitated lesion depth, and not affected by sound tissue thickness under the lesion. Clinical significance: Clinically, it is challenging to measure/scan at the same location and same angle longitudinally, however, it is important to standardize these parameters. Scanning within 20° deviation from perpendicular did not affect detection performance of PTR/LUM, and PTR/LUM value showed positive moderate correlation with caries depth.

Oriented Object Detection in Aerial Images Based on the Scaled Smooth L1 Loss Function

Although many state-of-the-art object detectors have been developed, detecting small and densely packed objects with complicated orientations in remote sensing aerial images remains challenging. For object detection in remote sensing aerial images, different scales, sizes, appearances, and orientations of objects from different categories could most likely enlarge the variance in the detection error. Undoubtedly, the variance in the detection error should have a non-negligible impact on the detection performance. Motivated by the above consideration, in this paper, we tackled this issue, so that we could improve the detection performance and reduce the impact of this variance on the detection performance as much as possible. By proposing a scaled smooth L1 loss function, we developed a new two-stage object detector for remote sensing aerial images, named Faster R-CNN-NeXt with RoI-Transformer. The proposed scaled smooth L1 loss function is used for bounding box regression and makes regression invariant to scale. This property ensures that the bounding box regression is more reliable in detecting small and densely packed objects with complicated orientations and backgrounds, leading to improved detection performance. To learn rotated bounding boxes and produce more accurate object locations, a RoI-Transformer module is employed. This is necessary because horizontal bounding boxes are inadequate for aerial image detection. The ResNeXt backbone is also adopted for the proposed object detector. Experimental results on two popular datasets, DOTA and HRSC2016, show that the variance in the detection error significantly affects detection performance. The proposed object detector is effective and robust, with the optimal scale factor for the scaled smooth L1 loss function being around 2.0. Compared to other promising two-stage oriented methods, our method achieves a mAP of 70.82 on DOTA, with an improvement of at least 1.26 and up to 16.49. On HRSC2016, our method achieves an mAP of 87.1, with an improvement of at least 0.9 and up to 1.4.

Vine copula Granger causality in quantiles

ABSTRACT Causal relationships between time series are typically examined by testing for Granger causality. Although many studies have tested Granger causality in mean, non-causality in mean does not need to carry over to other distribution characteristics or different parts of the distribution. This scenario has motivated many researchers to investigate causal relations from the perspective of conditional quantiles. Several methods have been proposed for both parametric and nonparametric modelling frameworks. Parametric methods have limitations in detecting nonlinear causality, whereas nonparametric methods have difficulty selecting smoothing parameters that significantly affect detection performance. To overcome the difficulties of both parametric and nonparametric Granger causality tests in quantiles, we propose a vine copula Granger causality test in quantiles using the semiparametric time-series modelling technique. The proposed test overcomes shortcomings in parametric modelling and has a computational advantage over nonparametric tests. Our test shows good performance in terms of size and power when using various simulated data. Finally, we illustrate our test using recent cryptocurrency data.

Confounds and overestimations in fake review detection: Experimentally controlling for product-ownership and data-origin.

The popularity of online shopping is steadily increasing. At the same time, fake product reviews are published widely and have the potential to affect consumer purchasing behavior. In response, previous work has developed automated methods utilizing natural language processing approaches to detect fake product reviews. However, studies vary considerably in how well they succeed in detecting deceptive reviews, and the reasons for such differences are unclear. A contributing factor may be the multitude of strategies used to collect data, introducing potential confounds which affect detection performance. Two possible confounds are data-origin (i.e., the dataset is composed of more than one source) and product ownership (i.e., reviews written by individuals who own or do not own the reviewed product). In the present study, we investigate the effect of both confounds for fake review detection. Using an experimental design, we manipulate data-origin, product ownership, review polarity, and veracity. Supervised learning analysis suggests that review veracity (60.26-69.87%) is somewhat detectable but reviews additionally confounded with product-ownership (66.19-74.17%), or with data-origin (84.44-86.94%) are easier to classify. Review veracity is most easily classified if confounded with product-ownership and data-origin combined (87.78-88.12%). These findings are moderated by review polarity. Overall, our findings suggest that detection accuracy may have been overestimated in previous studies, provide possible explanations as to why, and indicate how future studies might be designed to provide less biased estimates of detection accuracy.

MSSDet: Multi-Scale Ship-Detection Framework in Optical Remote-Sensing Images and New Benchmark

Ships comprise the only and most important ocean transportation mode. Thus, ship detection is one of the most critical technologies in ship monitoring, which plays an essential role in maintaining marine safety. Optical remote-sensing images contain rich color and texture information, which is beneficial to ship detection. However, few optical remote-sensing datasets are open publicly due to the issue of sensitive data and copyrights, and only the HRSC2016 dataset is built for the ship-detection task. Moreover, almost all general object detectors suffer from the failure of multi-scale ship detection because of the diversity of spatial resolution and ship size. In this paper, we re-annotate the HRSC2016 dataset and supplement 610 optical remote-sensing images to build a new open source ship-detection benchmark dataset with rich multi-scale ship objects named the HRSC2016-MS dataset. In addition, we further explore the potential of a recursive mechanism in the field of object detection and propose a novel multi-scale ship-detection framework (MSSDet) in optical remote-sensing images. The success of detecting multi-scale objects depends on the hierarchical pyramid structure in the object-detection framework. However, the inherent semantic and spatial gaps among hierarchical pyramid levels seriously affect detection performance. To alleviate this problem, we propose a joint recursive feature pyramid (JRFP), which can generate semantically strong and spatially refined multi-scale features. Extensive experiments were conducted on the HRSC2016-MS, HRSC2016, and DIOR datasets. Detailed ablation studies directly demonstrated the effectiveness of the proposed JRFP architecture and also showed that the proposed method has excellent generalizability. Comparisons with state-of-the-art methods showed that the proposed method achieves competitive performance, i.e., 77.3%, 95.8%, and 73.3% mean average precision accuracy on the HRSC2016-MS, HRSC2016, and DIOR datasets, respectively.

열변형을 고려한 함정 통합마스트 탑재 센서 성능 보정 방안에 관한 연구

In general, conformal type sensor’s alignment error is measured before and after mounting on the integrated mast and used for sensor’s performance compensation. However, in case of integrated mast made of metal, there is a possibility that additional deformation may occur due to external temperature and solar heat, and it could affect detection performance of conformal type radar even a slight deformation. In this paper, the thermal deformation level of the integrated mast considering external environment and detection performance error level of mounted sensor are described. Finally, a performance compensation method of the integrated mast mounted sensor is proposed. The thermal deformation level of the integrated mast is not dramatic considering its size, and detection performance error is calculated under 1%. However, those error should be compensated considering sensor’s detection range. Accordingly, performance compensation methods applicable to the integrated mast mounted sensor were proposed.

On the effect of sampling frequency on the electricity theft detection performance

Electricity theft is a common problem in many countries and affects the revenue of the electricity industry significantly. Recently, machine and deep learning techniques are being used widely to detect thieves by analysing the consumption patterns. While the prediction accuracy of these methods depends on the number and quality of the existing samples used for training models, the majority of previous research work focussed on data with high sampling frequencies, for example, data from smart grids. However, based on technological and operational limitations, the rate of metre reading is too low in many countries, which can affect detection performance. To investigate the effect of sampling frequency on the performance of detection methods, we designed a processing framework to evaluate different classification algorithms on versions of a challenging dataset obtained by down-sampling the original data at various rates. We used different features from time, frequency and time-frequency domains and their combinations in our investigations. Experimental results show that the performance of the models decreases with a small slope up to the sampling frequency of one data per week. The performance degradation rate is significant for data with lower sampling frequencies. We also proposed a bagging network that could improve the overall detection performance at different sampling frequencies.

Dynamic Anchor: A Feature-Guided Anchor Strategy for Object Detection

The majority of modern object detectors rely on a set of pre-defined anchor boxes, which enhances detection performance dramatically. Nevertheless, the pre-defined anchor strategy suffers some drawbacks, especially the complex hyper-parameters of anchors, seriously affecting detection performance. In this paper, we propose a feature-guided anchor generation method named dynamic anchor. Dynamic anchor mainly includes two structures: the anchor generator and the feature enhancement module. The anchor generator leverages semantic features to predict optimized anchor shapes at the locations where the objects are likely to exist in the feature maps; by converting the predicted shape maps into location offsets, the feature enhancement module uses the high-quality anchors to improve detection performance. Compared with the hand-designed anchor scheme, dynamic anchor discards all pre-defined boxes and avoids complex hyper-parameters. In addition, only one anchor box is predicted for each location, which dramatically reduces calculation. With ResNet-50 and ResNet-101 as the backbone of the one-stage detector RetinaNet, dynamic anchor achieved 2.1 AP and 1.0 AP gains, respectively. The proposed dynamic anchor strategy can be easily integrated into the anchor-based detectors to replace the traditional pre-defined anchor scheme.