Deep Edge Research Articles

Improved lane detection for autonomous vehicles using deep learning, semantic segmentation, edge detection and multi-sensor data fusion

Automated driving has gained significant attention because it can eliminate severe driving risks in real time. While autonomous vehicles rely heavily on sensors for lane detection, obstacle identification, and environmental awareness, accurate lane recognition remains a persistent challenge due to factors such as noise from shadows, poor lane markings, and obstructed views. Despite advances in computer vision, this problem is yet to be fully resolved, presenting a gap in the current literature. The primary objective of this research is to address these challenges by developing an enhanced lane-detection system. To achieve this, the study integrates advanced techniques, including semantic segmentation, edge detection, and deep learning, coupled with multi-sensor data fusion from cameras, LIDAR, and radar. By employing this methodology, the research examines various lane-detection methods and benchmarks the proposed model against existing systems in terms of accuracy, specificity and processing speed. Initial findings demonstrate that the combination of semantic segmentation and multi-sensor fusion improves lane detection in real-time scenarios. The proposed model achieved a lane detection accuracy of 97.8%, a specificity of 99.28%, and an average processing time of 0.0047 seconds per epoch. This study not only addresses the limitations of existing lane detection systems but also offers insights into improving road safety for autonomous vehicles.

Deep subwavelength topological edge state in a hyperbolic medium.

Topological photonics offers the opportunity to control light propagation in a way that is robust from fabrication disorders and imperfections. However, experimental demonstrations have remained on the order of the vacuum wavelength. Theoretical proposals have shown topological edge states that can propagate robustly while embracing deep subwavelength confinement that defies diffraction limits. Here we show the experimental proof of these deep subwavelength topological edge states by implementing periodic modulation of hyperbolic phonon polaritons within a van der Waals heterostructure composed of isotopically pure hexagonal boron nitride flakes on patterned gold films. The topological edge state is confined in a subdiffraction volume of 0.021 µm3, which is four orders of magnitude smaller than the free-space excitation wavelength volume used to probe the system, while maintaining the resonance quality factor above 100. This finding can be directly extended to and hybridized with other van der Waals materials to broadened operational frequency ranges, streamline integration of diverse polaritonic materials, and compatibility with electronic and excitonic systems.

An efficient water quality index forecasting and categorization using optimized Deep Capsule Crystal Edge Graph neural network.

The world's freshwater supply, predominantly sourced from rivers, faces significant contamination from various economic activities, confirming that the quality of river water is critical for public health, environmental sustainability, and effective pollution control. This research addresses the urgent need for accurate and reliable water quality monitoring by introducing a novel method for estimating the water quality index (WQI). The proposed approach combines cutting-edge optimization techniques with Deep Capsule Crystal Edge Graph neural networks, marking a significant advancement in the field. The innovation lies in the integration of a Hybrid Crested Porcupine Genghis Khan Shark Optimization Algorithm for precise feature selection, ensuring that the most relevant indicators of water quality (WQ) are utilized. Furthermore, the use of the Greylag Goose Optimization Algorithm to fine-tune the neural network's weight parameters enhances the model's predictive accuracy. This dual optimization framework significantly improves WQI prediction, achieving a remarkable mean squared error (MSE) of 6.7 and an accuracy of 99%. By providing a robust and highly accurate method for WQ assessment, this research offers a powerful tool for environmental authorities to proactively manage river WQ, prevent pollution, and evaluate the success of restoration efforts. PRACTITIONER POINTS: Novel method combines optimization and Deep Capsule Crystal Edge Graph for WQI estimation. Preprocessing includes data cleanup and feature selection using advanced algorithms. Deep Capsule Crystal Edge Graph neural network predicts WQI with high accuracy. Greylag Goose Optimization fine-tunes network parameters for precise forecasts. Proposed method achieves low MSE of 6.7 and high accuracy of 99%.

Brain Magnetic Resonance Image Inpainting via Deep Edge Region-based Generative Adversarial Network

Brain Magnetic Resonance Image Inpainting via Deep Edge Region-based Generative Adversarial Network

Enhancing Aviation Safety through AI-Driven Mental Health Management for Pilots and Air Traffic Controllers.

This article provides an overview of the mental health challenges faced by pilots and air traffic controllers (ATCs), whose stressful professional lives may negatively impact global flight safety and security. The adverse effects of mental health disorders on their flight performance pose a particular safety risk, especially in sudden unexpected startle situations. Therefore, the early detection, prediction and prevention of mental health deterioration in pilots and ATCs, particularly among those at high risk, are crucial to minimize potential air crash incidents caused by human factors. Recent research in artificial intelligence (AI) demonstrates the potential of machine and deep learning, edge and cloud computing, virtual reality and wearable multimodal physiological sensors for monitoring and predicting mental health disorders. Longitudinal monitoring and analysis of pilots' and ATCs physiological, cognitive and behavioral states could help predict individuals at risk of undisclosed or emerging mental health disorders. Utilizing AI tools and methodologies to identify and select these individuals for preventive mental health training and interventions could be a promising and effective approach to preventing potential air crash accidents attributed to human factors and related mental health problems. Based on these insights, the article advocates for the design of a multidisciplinary mental healthcare ecosystem in modern aviation using AI tools and technologies, to foster more efficient and effective mental health management, thereby enhancing flight safety and security standards. This proposed ecosystem requires the collaboration of multidisciplinary experts, including psychologists, neuroscientists, physiologists, psychiatrists, etc. to address these challenges in modern aviation.

Boosting edge detection via Fusing Spatial and Frequency Domains

Deep learning-based edge detection methods have shown great advantages and obtained promising performance. However, most of the current methods only extract features from the spatial (RGB) domain for edge detection and the information that can be mined is limited. As a result, they could not work well for the scenarios where the object is similar in color to the background. To combat this challenge, we propose a novel edge detection method by incorporating the features of both spatial and frequency domains, named Fusing Spatial and Frequency Domains (FSFD). A Frequency Perception (FP) module is constructed to extract the edges of objects in the frequency domain, which can avoid the indistinguishable situation in the spatial domain due to similar colors. A Multi-Scale Enhancement (MSE) module is designed to learn multi-scale feature in spatial domain, enabling the model to perceive edges of small objects. Spatial-Frequency Fusion (S2F) module is further introduced to fuse the features of spatial and frequency domains using an online learning manner. Adequate experiments are conducted on popular BSDS500, NYUDV2, and Multicue datasets. The results show that our method can outperform other methods as the state-of-the-art when dealing with the problem of edge detection. The codes will be released on https://github.com/JingDongdong/FSFD.

Remote Sensing Image Classification Based on Canny Operator Enhanced Edge Features.

Remote sensing image classification plays a crucial role in the field of remote sensing interpretation. With the exponential growth of multi-source remote sensing data, accurately extracting target features and comprehending target attributes from complex images significantly impacts classification accuracy. To address these challenges, we propose a Canny edge-enhanced multi-level attention feature fusion network (CAF) for remote sensing image classification. The original image is specifically inputted into a convolutional network for the extraction of global features, while increasing the depth of the convolutional layer facilitates feature extraction at various levels. Additionally, to emphasize detailed target features, we employ the Canny operator for edge information extraction and utilize a convolution layer to capture deep edge features. Finally, by leveraging the Attentional Feature Fusion (AFF) network, we fuse global and detailed features to obtain more discriminative representations for scene classification tasks. The performance of our proposed method (CAF) is evaluated through experiments conducted across three openly accessible datasets for classifying scenes in remote sensing images: NWPU-RESISC45, UCM, and MSTAR. The experimental findings indicate that our approach based on incorporating edge detail information outperforms methods relying solely on global feature-based classifications.

Intelligent Infrastructure for Traffic Monitoring Based on Deep Learning and Edge Computing

In the field of traffic management and control systems, we are witnessing a symbiotic evolution, where intelligent infrastructure is progressively collaborating with smart vehicles to produce benefits for traffic monitoring and security, by rapidly identifying hazardous behaviours. This exponential growth is due to the rapid development of deep learning in recent years, as well as the improvements in computer vision models. These technologies allow for monitoring tasks without the need to install numerous sensors or stop the traffic, using the extensive camera network of surveillance cameras already present in worldwide roads. This study proposes a computer vision-based solution that allows for real-time processing of video streams through edge computing devices, eliminating the need for Internet connectivity or dedicated sensors. The proposed system employs deep learning algorithms and vision techniques that perform vehicle detection, classification, tracking, speed estimation, and vehicle geolocation.

LVP-net: A deep network of learning visual pathway for edge detection

Deep learning-based edge detectors typically consist of the encoder and the decoder. To integrate multi-scale features into a global edge map effectively, researchers utilize classification networks such as VGG16 as the encoder and focus on the decoder architecture. In contrast to existing approaches, we propose a novel deep network for edge detection called learning-visual-pathway network (LVP-Net), in which an enhancer-encoder-decoder architecture is designed inspired by the biological visual pathway: the retina/lateral geniculate nucleus→the primary visual cortex (V1) → V2 → V4 → the inferior temporal cortex (IT). To simulate the visual mechanisms along this pathway, we design a feature enhancer network (FENet) that boosts the feature representation capability of the encoder. FENet is combined with VGG16 based on the hierarchical structure of the pathway. Furthermore, inspired by the integration ability of multiple features in IT, we introduce a feedforward fusion module (FFM). Finally, we evaluate LVP-Net on three benchmark datasets, i.e., BSDS500, NYUDv2, and Multicue. Experimental results demonstrate that our method achieves competitive performance compared with most state-of-the-art approaches.

Oncologic Outcomes for Squamous Cell Carcinoma In Situ With a Clinically Resolved Biopsy Site Managed by Watchful Waiting.

Treatment option decisions for low-risk squamous cell carcinoma in situ (SCCIS) are hampered by a paucity of management-type-specific outcomes data. Describe SCCIS tumor outcomes managed by watchful waiting and risk factors associated with poor cancer outcomes. Retrospective cohort study. Setting: Single academic hospital in a rural setting. Patients: Adults with SCCIS diagnosed between January 01, 2014, and December 31, 2016. Main Outcomes and Measures: Hazard ratios (HRs) for local recurrence (LR), nodal metastases (NM), distant metastases (DM), and disease-specific death (DSD). A total of 411 consecutive SCCIS tumors that were considered clinically resolved at follow-up and managed with watchful waiting were included. Seventeen tumors recurred locally. No instances of NM, DM, or DSD were identified. Multivariate analysis found that solid-organ transplant recipient status conferred the highest risk of local recurrence [HR, 9.979 (95% CI, 2.249-39.69)]. Additional risk factors predicting LR include anatomic location on the vermilion lip or ear [HR, 9.744 (95% CI, 1.420-69.28)], anatomic location on the head and neck [HR, 6.687 (95% CI, 1.583-36.15)], and a biopsy with tumor extending to the deep edge [HR, 6.562 (95% CI, 1.367-39.04)]. Watchful waiting for SCCIS with a clinically resolved biopsy site has a local recurrence rate of 4%.

Real-time railroad track components inspection framework based on YOLO-NAS and edge computing

The demand for efficient track inspection systems in the rapidly evolving rail transportation field is more pronounced than ever. Hence, this study combines deep learning and edge computing for railroad track component inspection, focusing on the YOLO-NAS architecture. Our objective was twofold: to harness the advantages of YOLO-NAS for accurate and high-speed detection while addressing the computational constraints of edge devices. Consequently, YOLO-NAS-S-PTQ model achieved a remarkable balance, with 74.77% mAP and 92.20 FPS, on the NVIDIA Jetson Orin platform. By deploying this model on an edge device and utilizing a multiprocessor pipeline, we observed an inference speed of 60.468 FPS, which was nearly double the rate of its single-threaded counterpart. Field tests further confirmed the efficiency of the model, demonstrating a recall rate of 80.77% and an accuracy of 96.64%. These findings underscore the potential of YOLO-NAS in transforming traditional rail component inspection methods, significantly reducing human interventions and potential errors.

TemPanSharpening: A multi-temporal Pansharpening solution based on deep learning and edge extraction

The tradeoff among spatial, temporal, and spectral resolution of remote sensing (RS) images due to sensor properties limits the development of RS applications. Most image enhancement studies tend to focus on either spatio-temporal fusion or spatio-spectral fusion. As a more comprehensive solution, spatial–temporal-spectral fusion (STSF) is complicated but its potential is worth to be further explored. In this study, we propose a novel STSF method from the perspective of multi-temporal Pansharpening. Canny edge extraction is applied to Panchromatic (PAN) images to identify edges while avoiding the disruption of multi-temporal land cover changes. We then build a TemPanSharpening net (TPSnet) which only uses one high-spatio-low-spectra-temporal PAN and one low-spatio-high-spectra-temporal multispectral image as input. TPSnet follows a super-resolution structure and embeds two basic modules: residual-in-residual dense blocks (RRDB) and convolutional block attention module (CBAM). A series of interior ablation experiments were conducted on TPSnet and we also compared it with some representative spatio-temporal fusion, Pansharpening, and STSF algorithms. TPSnet presented satisfactory performance on complicated meter-level ground surfaces according to the quantitative evaluation result, and it demonstrated excellent robustness to land cover change.

A congruent-melting rubidium zinc vanadium oxychloride exhibiting strong mid-infrared second-harmonic generation and high laser-induced damage threshold

A novel mid-infrared (MIR) nonlinear optical (NLO) crystal Rb4ZnV5O15Cl (RZVC) has been designed and sythesized by the aliovalent substitution strategy based on the parent structure Rb2(VO)(V2O7). Modification of square pyramidal VO5 units through introduction of chlorine ion and zinc metal caion significantly enhanced the dipole moments in a unit cell from 1.653 D for Rb2(VO)(V2O7) to 6.559 D for RZVC. Consequently, polycrystalline RZVC exhibits strong SHG responses of 12.5 × KDP @1064 nm and 1.2 × AGS @2 μm. Furthermore, RZVC features a congruent melting property, an ultra-high LIDT value (62 × AGS @1064 nm), and a deep IR transparency cutoff edge (10 μm). Most importantly, RZVC can achieve full-wavelength phase matching in a very wide MIR transparency range of 1.69–10 μm, completely covering the important atmospheric window (3–5 μm) through NLO processes. Because of its many outstanding properties, RZVC can be a highly desirable candidate for practical MIR NLO applications.

Edge AI Model Deployed for Real-Time Detection of Atrial Fibrillation Risk during Sinus Rhythm.

Objectives: The study aimed to develop a deep learning-based edge AI model deployed on electrocardiograph (ECG) devices for the real-time detection of atrial fibrillation (AF) risk during sinus rhythm (SR) using standard 10 s, 12-lead electrocardiograms (ECGs). Methods: A novel approach was used to convert standard 12-lead ECGs into binary images for model input, and a lightweight convolutional neural network (CNN)-based model was trained using data collected by the Japan Agency for Medical and Research Development (AMED) between 2019 and 2022. Patients over 40 years old with digital, SR ECGs were retrospectively enrolled and divided into AF and non-AF groups. The data labeling was supervised by cardiologists. The dataset was randomly allocated into training, validation, and internal testing datasets. External testing was conducted on data collected from other hospitals. Results: The best-trained model achieved an AUC of 0.82 and 0.80, sensitivity of 79.5% and 72.3%, specificity of 77.8% and 77.7%, precision of 78.2% and 76.4%, and overall accuracy of 78.6% and 75.0% in the internal and external testing datasets, respectively. The deployed model and app package utilized 2.5 MB and 40 MB of the available ROM and RAM capacity on the edge ECG device, correspondingly. The processing time for AF risk detection was approximately 2 s. Conclusions: The model maintains comparable performance and improves its suitability for deployment on resource-constrained ECG devices, thereby expanding its potential impact to a wide range of healthcare settings. Its successful deployment enables real-time AF risk detection during SR, allowing for timely intervention to prevent AF-related serious consequences like stroke and premature death.

A NOVEL STUDY ON MACHINE LEARNING ALGORITHMS FOR BIG DATA

Big data's vastness and complexity pose a formidable challenge to traditional data analysis methods. Machine learning algorithms emerge as intrepid navigators, extracting meaningful patterns and hidden correlations from the deluge of information. Their versatility handles heterogeneous data formats, while their robust mechanisms ensure data quality. Machine learning empowers predictive modeling, anomaly detection, recommendation systems, fraud detection, and customer segmentation. Implementing these algorithms in big data environments presents challenges in data quality, scalability, and interpretability. Emerging trends like deep learning, edge computing, and explainable AI offer promising solutions, paving the way for a future where big data and machine learning shape data-driven decision-making. Keywords: Machine Learning, Data Quality, Recommendation system, deep learning.

Deep Learning-Enabled Joint Edge Content Caching and Power Allocation Strategy in Wireless Networks

Deep Learning-Enabled Joint Edge Content Caching and Power Allocation Strategy in Wireless Networks

Enhancing Deep Edge Detection through Normalized Hadamard-Product Fusion.

Deep edge detection is challenging, especially with the existing methods, like HED (holistic edge detection). These methods combine multiple feature side outputs (SOs) to create the final edge map, but they neglect diverse edge importance within one output. This creates a problem: to include desired edges, unwanted noise must also be accepted. As a result, the output often has increased noise or thick edges, ignoring important boundaries. To address this, we propose a new approach called the normalized Hadamard-product (NHP) operation-based deep network for edge detection. By multiplying the side outputs from the backbone network, the Hadamard-product operation encourages agreement among features across different scales while suppressing disagreed weak signals. This method produces additional Mutually Agreed Salient Edge (MASE) maps to enrich the hierarchical level of side outputs without adding complexity. Our experiments demonstrate that the NHP operation significantly improves performance, e.g., an ODS score reaching 0.818 on BSDS500, outperforming human performance (0.803), achieving state-of-the-art results in deep edge detection.

Rapid sea level rise causes loss of seagrass meadows

As global declines in seagrass populations continue to cause great concern, long-term assessment of seagrass meadows show promise in furnishing valuable clues into fundamental causes of seagrass loss and drivers of environmental change. Here we report two long-term records of seagrass presence in western Gulf of Mexico coastal waters (Laguna Madre) that provided insight into their rapid decline in a relatively pristine ecosystem. Coincident with unprecedented increases in water depth starting in 2014 (14–25 mm y−1), monthly measurements at a deep edge fixed station revealed that two ubiquitous seagrass species (Halodule wrightii and Syringodium filiforme) vanished altogether in just five years; a subsequent basin-wide assessment revealed that seagrasses disappeared at 23% of 144 sentinel stations. Models that incorporate differing sea level rise scenarios and water depth thresholds reveal potential global losses of seagrass habitat (14,000 km2), with expansion into newly created shallow habitats constrained by altered natural shorelines.

Evaluation of alternative approaches to PHABSIM modeling of coastal cutthroat trout spawning habitat

AbstractIn the face of a changing climate and increasing human demand for water, an understanding of habitat preference has become critical for managing wild fish populations and projecting potential changes in habitat and populations. Two approaches to Physical Habitat Simulation (PHABSIM) modeling of coastal cutthroat trout Oncorhynchus clarkii clarkii spawning habitat were compared by modeling a reach, consisting of eight transects covering a pool and upstream and downstream riffles, of a small Puget Sound stream with two sets of habitat variables. The reach contained a cluster of redds 2 years in a row, assumed to be an indication of preferred spawning habitat and was located in the area of maximum spawning in the watershed, based on multiple years of redd surveys. Two PHABSIM instream flow models of the study site, one based on standard microhabitat (depth [D], velocity [V], and substrate [S]) suitabilities and the other based on D, V, and channel unit (CU) suitabilities, were developed and compared for their relative ability to correctly predict coastal cutthroat trout spawning habitat selection at the redd cluster within the PHABSIM study site. One approach was the standard use of habitat suitability criteria (HSC) for D, V, and S to indicate spawning habitat quality. The alternate approach was to replace substrate HSC with CU index HSC that incorporated dominant substrate particle diameter, CU (riffle, deep and shallow pool tail, pool body, deep and shallow pool edge, cascade, waterfall, and terrestrial), where deep and shallow units were based on relative residual depth (RRD), size of CU relative to channel width, and position within CU. Spawning habitat quality was calculated for each transect as weighted usable width with the standard HSC metrics (WUWs) as well as the modified CU index (WUWm). WUWm at the transects bracketing the redd cluster exceeded WUWm at the remaining six transects and was outside the 95% confidence interval for WUWm at the remaining transects. In contrast, WUWs varied less between the redd cluster and the remainder of the transects, suggesting the CU index better reflected spawning habitat quality than substrate. Incorporation of elevation relative to SZF addressed vulnerability to declining flow during incubation. Both models resulted in maximum WUA within the range of discharges that coincided with the majority of fresh cutthroat trout redds in Skookum Creek.

Automatic vectorization of historical maps: A benchmark.

Shape vectorization is a key stage of the digitization of large-scale historical maps, especially city maps that exhibit complex and valuable details. Having access to digitized buildings, building blocks, street networks and other geographic content opens numerous new approaches for historical studies such as change tracking, morphological analysis and density estimations. In the context of the digitization of Paris atlases created in the 19th and early 20th centuries, we have designed a supervised pipeline that reliably extract closed shapes from historical maps. This pipeline is based on a supervised edge filtering stage using deep filters, and a closed shape extraction stage using a watershed transform. It relies on probable multiple suboptimal methodological choices that hamper the vectorization performances in terms of accuracy and completeness. Objectively investigating which solutions are the most adequate among the numerous possibilities is comprehensively addressed in this paper. The following contributions are subsequently introduced: (i) we propose an improved training protocol for map digitization; (ii) we introduce a joint optimization of the edge detection and shape extraction stages; (iii) we compare the performance of state-of-the-art deep edge filters with topology-preserving loss functions, including vision transformers; (iv) we evaluate the end-to-end deep learnable watershed against Meyer watershed. We subsequently design the critical path for a fully automatic extraction of key elements of historical maps. All the data, code, benchmark results are freely available at https://github.com/soduco/Benchmark_historical_map_vectorization.