Manual analysis and diagnosis of COVID-19 through the examination of Computed Tomography (CT) images of the lungs can be time-consuming and result in errors, especially given high volume of patients and numerous image...
详细信息
ISBN:
(纸本)9798350365474
Manual analysis and diagnosis of COVID-19 through the examination of Computed Tomography (CT) images of the lungs can be time-consuming and result in errors, especially given high volume of patients and numerous images per patient. So, we address the need for automation of this task by developing a new deep learning-based pipeline. Our motivation was sparked by the CVPR Workshop on "Domain Adaptation, Explainability and Fairness in AI for Medical Image Analysis", more specifically, the "COVID-19 Diagnosis Competition (DEF-AI-MIA COV19D)" under the same Workshop. this challenge provides an opportunity to assess our proposed pipeline for COVID-19 detection from CT scan images. the same pipeline incorporates one of the architectures in the EfficientNet "family", but with an added Spatial Attention Mechanism: EfficientNet-SAM. Also, unlike the traditional/past pipelines, which relied on a preprocessing step, our pipeline takes the raw selected input images without any such step, except for an image-selection step to simply reduce the number of CT images required for training and/or testing. Moreover, our pipeline is computationally efficient, as, for example, it does not incorporate a decoder for segmenting the lungs. It also does not combine different models nor combine RNN with a backbone, as other pipelines in the past did. Nevertheless, our pipeline outperformed all approaches presented by other teams in last year's instance of the same challenge using the validation subset. It also placed 5th in this year's competition, ranking less than 1.3% below the 1st place and close to 3.5% above the 6th place based on the macro-F1 score.
Bitter gourd (Momordica charantia) is an essential ingredient used across diverse industries, valued not only for culinary purposes but also recognized for its various health and medical benefits. this vine is known f...
详细信息
In order to achieve the process conditions of phase-controlled ultrasonic nondestructive testing, a robotassisted trajectory planning method of curved workpieces for NDT is proposed in this paper. Combined with comput...
详细信息
ISBN:
(纸本)9798350311259
In order to achieve the process conditions of phase-controlled ultrasonic nondestructive testing, a robotassisted trajectory planning method of curved workpieces for NDT is proposed in this paper. Combined withcomputervision and robotics, tasks such as 3D scene reconstruction, target recognition and path pose planning are implemented and fully autonomous NDT is achieved. Firstly, an environment based on 3D vision is built, and the scene point cloud is obtained through 3D reconstruction technology. then the point cloud model of the workpiece to be detected is matched withthe real-time scene point cloud, and the Euclidean clustering and SHOT feature descriptor are used to separate the scene. Finally, based on the recognized target point cloud, the robot detection trajectory is generated by plane slicing. the simulation results show that the point cloud model to be inspected obtained through segmentation and recognition can entirely and effectively express the entire workpiece to be reviewed. the generated trajectory points meet the accuracy requirements and pose requirements of NDT.
Image defogging is an important computervision method. the current end-to-end defogging method, Convolutional Neural Networks (CNNs), has achieved significant success, and estimating depth information has become a ma...
详细信息
ISBN:
(纸本)9798400708039
Image defogging is an important computervision method. the current end-to-end defogging method, Convolutional Neural Networks (CNNs), has achieved significant success, and estimating depth information has become a mainstream method. However, depth information only focuses on the type of object, resulting in texture information loss and edge blur issues. To solve this *** article introduces a new defogging network (PHD) based on physical and image reconstruction. We first use a depth estimation network to obtain depth information and learn to physically decompose haze images into two components that conform to the scattering model: transmitted images and atmospheric light. In addition, we combine depth information with prior details of the fog map, learn the areas we need to focus on through the designed attention network, and generate weight matrix parameters to guide defogging. In the fine-tuning stage, we use the image hyper variational algorithm to repair its texture details. the experimental results indicate that compared with existing algorithms, this method has good performance.
To effectively improve the robot's ability to perceive the external environment. this paper proposes a visual-tactile fusion object attribute classification method based on the GAMP attention mechanism. the method...
详细信息
To address the issues of low intelligence level and inefficient inspection of high-voltage transmission line patrols, this study combines computervision technology to propose an improved YOLOv5-based detection algori...
详细信息
ISBN:
(纸本)9798400708039
To address the issues of low intelligence level and inefficient inspection of high-voltage transmission line patrols, this study combines computervision technology to propose an improved YOLOv5-based detection algorithm, named YOLOv5_ACG, for detecting bird nests on high-voltage transmission lines. the algorithm addresses challenges such as significant scale variations, indistinct features, and the potential for false positives and false ***, a feature fusion mechanism is employed in the neck network to optimize the detection head, improving the feature fusion to handle the diverse scales of targets. Secondly, an enhanced learnable parallel weighted attention module is utilized to enhance the model's ability to focus on bird nest features. Lastly, lightweight convolutions are employed to reduce network parameters and computations, minimizing real-time performance losses resulting from increased model *** results demonstrate that YOLOv5_ACG achieves a 93.0% mAP@0.5 on a self-made transmission line bird nest dataset, which is a 2.2% improvement over YOLOv5. this algorithm enhances the detection accuracy of bird nests on transmission lines in practical application scenarios, making it more suitable for embedded devices used in actual transmission line bird nest detection applications.
Recent trends of industry brought to the attention of the scientific and practitioners' communities the need for providing efficient strategies to manage the End of Life of products. this propensity gained increas...
详细信息
Aiming at the difficulty and slow speed of multi-pose insulator pin detection in single-chip insulator replacement robot in substation insulator replacement robot, a fast and reliable insulator pin detection method YO...
详细信息
Medical image segmentation plays an essential role in developing computer-assisted diagnosis and treatment systems, yet it still faces numerous challenges. In the past few years, Convolutional Neural Networks (CNNs) h...
详细信息
ISBN:
(纸本)9789819985425;9789819985432
Medical image segmentation plays an essential role in developing computer-assisted diagnosis and treatment systems, yet it still faces numerous challenges. In the past few years, Convolutional Neural Networks (CNNs) have been successfully applied to the task of medical image segmentation. Regrettably, due to the locality of convolution operations, these CNN-based architectures have their limitations in learning global context information in images, which might be crucial to the success of medical image segmentation. Meanwhile, the vision Transformer (ViT) architectures own the remarkable ability to extract long-range semantic features withthe shortcoming of their computation complexity. To make medical image segmentation more efficient and accurate, we present a novel light-weight architecture named LeViT-UNet, which integrates multi-stage Transformer blocks in the encoder via LeViT, aiming to explore the effectiveness of fusion between local and global features together. Our experiments on two challenging segmentation benchmarks indicate that the proposed LeViT-UNet achieved competitive performance compared with various state-of-the-art methods in terms of efficiency and accuracy, suggesting that LeViT can be a faster feature encoder for medical images segmentation. LeViT-UNet-384, for instance, achieves Dice similarity coefficient (DSC) of 78.53% and 90.32% with a segmentation speed of 85 frames per second (FPS) in the Synapse and ACDC datasets, respectively. therefore, the proposed architecture could be beneficial for prospective clinic trials conducted by the radiologists. Our source codes are publicly available at https://***/apple1986/LeViT_UNet.
the deep learning-based methods have shown promising performance in restoring degraded images such as underwater and haze images. However, the majority of existing methods rely on simplified imaging models, which limi...
详细信息
ISBN:
(纸本)9789819985517;9789819985524
the deep learning-based methods have shown promising performance in restoring degraded images such as underwater and haze images. However, the majority of existing methods rely on simplified imaging models, which limits their generalization and applicability in real-world scenarios. To address these issues, we incorporate the imaging mechanism in complex underwater environments to redefine the imaging model for degraded images under medium propagation. We then propose a multi-stage restoration framework that combines model-based iterative optimization methods and deep learning methods. At the same time, to tackle the problem of inaccurate parameter estimation in methods relying on a single prior, we introduce a regularization design based on joint priors and develop an attention-based color correction network to correct color distortions in the degraded images. Experimental results on real-world degraded images demonstrate the effectiveness and superiority of our method in both quantitative and subjective evaluations when compared to state-of-the-art methods.
暂无评论