Automating the diagnosing process has shown promising results in recent years due to the advancements in deep learning approaches. In this work, we provide a unique skin disease diagnosis system that uses Convolutiona...
详细信息
The diagnosis of a range of eye disorders needs to categorize the retinal vessels. Computerized implementation of this process is becoming increasingly essential for automated screening systems for retinal diseases. T...
详细信息
ISBN:
(数字)9798350356816
ISBN:
(纸本)9798350356823
The diagnosis of a range of eye disorders needs to categorize the retinal vessels. Computerized implementation of this process is becoming increasingly essential for automated screening systems for retinal diseases. To achieve a more accurate extraction of the retinal vessels, a new pre-processing step is proposed. These proposed pre-processes are also compared to other algorithms to assess their impact. The proposed pre-processing process consists of two phases. The first phase is the implementation and validation of the pre-processing modules, and the second phase is the implementation of these pre-processes onto the retinal vessels that were to be extracted. To achieve a significantly improved segmented vessel image, the proposed pre-process phase employs a common image-processing technique. In recent years, there has been a great deal of focus on retinal vessel identification studies, and the importance of assessing and confirming the findings of retinal vessel segmentation.
a kind of spatial-temporal neural network video smoke detection algorithm is proposed in order to solve the problems associated with the incorrect classification of the static approximate smoke background in the face ...
详细信息
ISBN:
(数字)9798350376548
ISBN:
(纸本)9798350376555
a kind of spatial-temporal neural network video smoke detection algorithm is proposed in order to solve the problems associated with the incorrect classification of the static approximate smoke background in the face of the detection of smoke in video detection networks, and the problem of false alarms and of the original test model algorithms being different in different detection environments. Based on the original YOLO v4 neural network algorithm, this paper introduces a k-means + + algorithm and genetic algorithm, while using the algorithm's clustering function to classify the sample points of the real boxes of the image data set, which make it a more suitable anchor. At the same time, the genetic algorithm is used to adjust its anchor in order to allow the generated anchor to adapt to the needs related to smoke detection. In the original neural network model, the dual-stream network model algorithm is used to extract information from the first step of the YOLO algorithm in order to further filter the smoke's characteristics as well as filter out error information, all to improve the detection capabilities of the overall neural network for video smoke fog images. Compared with traditional YOLOv4 networks, the algorithm obtained by the model algorithm has been improved by 8.51°/0. In actual tests, the alarm time requirements of the smoke alarm test program for early fire monitoring and the alarm systems for visual images were improved, and the detection accuracy of the network was also improved based on the assurance of the detection speed, while the performance of the model algorithm was also improved for different scenes.
Studying circumstellar environments is crucial for understanding exoplanets and stellar systems. Instruments like SPHERE can extract information about these environments by leveraging advanced image reconstruction met...
详细信息
ISBN:
(数字)9789464593617
ISBN:
(纸本)9798331519773
Studying circumstellar environments is crucial for understanding exoplanets and stellar systems. Instruments like SPHERE can extract information about these environments by leveraging advanced image reconstruction methods, possibly based on deep learning. This work focuses on unfolded proximal neural networks based on Condat- vii iterations and proposes a new nonlinear formulation. To evaluate and compare the performance of the proposed reconstruction strategies, two datasets dedicated to circumstellar environments analysis in the context of high-contrast imagery have been created offering different level of complexity in the evaluation of the performance.
Pneumonia is an infectious disease characterized by inflammation of the lungs' air sacs, which results in the accumulation of fluid or pus. Medical images is important for the timely identification and precise dia...
详细信息
Pneumonia is an infectious disease characterized by inflammation of the lungs' air sacs, which results in the accumulation of fluid or pus. Medical images is important for the timely identification and precise diagnosis of illnesses;chest X-rays are a commonly utilized modality for respiratory disorders including pneumonia. In this research, optimized double transformer residual super-resolution network-related chest x-ray imageries for the classification of pneumonia identification (DTRSN-XRI-CPI). The procedure involves pre-processing the input image using region-aware neural graph collaborative filtering (RNGCF) to reduce noise, enhance contrast, and eliminate high and low frequencies from the collected dataset. Next, the Synchro-squeezed fractional wavelet transform (SFWT) is utilized for the feature extraction to extract color features such as color, shape, spatial, texture, and relation from the image. Hence, the weight parameters for DTRSN are optimized using the Hunter Prey Optimization algorithms (HPOA). Then the DTRSN-XRI-CPI is implemented in Python and the performance metrics like precision, accuracy, recall, specificity, F1-score, and ROC are analysed. The performance of the DTRSN-XRI-CPI approach attains 20.7 %, 22.6 % and 30.5 % higher accuracy;21.8 %, 29.3 % and 30.5 %higher precision and 21.8 %, 29.5 % and 32.6 % higher recall when analysed through existing an intelligent computational framework based on deep learning for the identification and classification of pneumonia illness (ICPDDL-ICF), an adaptive and altruistic deep feature selection approach based on PSO for pneumonia detection from chest X-rays (APSO-DFSM-PDCX) and a deep learning system that uses explainable AI (DLAIB-PI-EAI) techniques respectively.
image classification is one of the most fundamental capabilities of machine vision intelligence. In this work, we revisit the image classification task using visually-grounded language models (vLMs) such as GPT-4v and...
ISBN:
(纸本)9798331314385
image classification is one of the most fundamental capabilities of machine vision intelligence. In this work, we revisit the image classification task using visually-grounded language models (vLMs) such as GPT-4v and LLavA. We find that existing proprietary and public vLMs, despite often using CLIP as a vision encoder and having many more parameters, significantly underperform CLIP on standard image classification benchmarks like imageNet. To understand the reason, we explore several hypotheses concerning the inference algorithms, training objectives, and data processing in vLMs. Our analysis reveals that the primary cause is data-related: critical information for image classification is encoded in the vLM's latent space but can only be effectively decoded with enough training data. Specifically, there is a strong correlation between the frequency of class exposure during vLM training and instruction-tuning and the vLM's performance in those classes; when trained with sufficient data, vLMs can match the accuracy of state-of-the-art classification models. Based on these findings, we enhance a vLM by integrating classification-focused datasets into its training, and demonstrate that the enhanced classification performance of the vLM transfers to its general capabilities, resulting in an improvement of 11.8% on the newly collected imageWikiQA dataset. Project page: https://***/vLMClassifier-Website/.
Counterfeiting poses a significant threat as the circulation of fake currency diminishes the value of genuine notes, thereby disrupting the country's economic stability. Addressing this issue is critical before th...
详细信息
ISBN:
(数字)9798331530013
ISBN:
(纸本)9798331530020
Counterfeiting poses a significant threat as the circulation of fake currency diminishes the value of genuine notes, thereby disrupting the country's economic stability. Addressing this issue is critical before the proliferation of counterfeit notes becomes unmanageable. Manual detection methods are often unreliable since counterfeit notes are produced with materials and inks that closely resemble the original. Across the country, counterfeit detection is typically carried out using hardware-based systems. However, these methods are time-consuming and struggle to process large volumes efficiently. To address these challenges and streamline the process of counterfeit currency detection, this study proposes an imageprocessing-based computational technique. The objective is to accurately determine whether a given note is genuine or counterfeit, with a high prediction rate. This detection is enhanced using deep learning algorithms, which analyze key attributes such as color, form, paper thickness, serial numbers, and image filters on the currency. The proposed model is trained on a real-time dataset consisting of both genuine and counterfeit notes. Experimental results demonstrate that this method achieves an overall accuracy of 96.41%.
Detection of corrosion in moving objects like ships is challenging due to the dynamic nature of the input image. Existing machine learning techniques are suitable for static images and the algorithms suffer in perform...
详细信息
ISBN:
(数字)9798350368109
ISBN:
(纸本)9798350368116
Detection of corrosion in moving objects like ships is challenging due to the dynamic nature of the input image. Existing machine learning techniques are suitable for static images and the algorithms suffer in performance when is a live video. In this paper, imageprocessing for detecting corrosion using YOLOv8 which more suitable for processing live videos as speed and accuracy is better. This makes YOLO v8 for corrosion detection in live videos. In addition, Weights and Biases (W&B). is used in the algorithm as it is pivotal in establishing the connections between neurons and biases helps in circumventing flexible inputs. By combining YOLOv8's and W&B approach the accuracy and efficiency of corrosion detection systems is improved. This can ultimately assist in better maintenance and preservation of essential infrastructure resources.
Conventional ISP pipelines and image enhancement methods are designed and optimized for human vision, creating a gap between the requirements of computer and human visions. To bridge the requirement gap, we present a ...
详细信息
ISBN:
(数字)9798350330991
ISBN:
(纸本)9798350331004
Conventional ISP pipelines and image enhancement methods are designed and optimized for human vision, creating a gap between the requirements of computer and human visions. To bridge the requirement gap, we present a co-design framework in which backend computer vision plays a pivotal role in shaping the proceeding imageprocessing algorithm. It features a pre- processing adapter network, responsible for the restoration and enhancement of RAW images from computer vision perspective, especially in challenging environmental conditions. Specifically, we extract feature maps from the backend vision network, utilizing them as constraints for optimizing the preprocessing adapter network. To validate the effectiveness of our proposed framework, we employ object detection in low-light conditions as the computer vision task, with YOLO-v5 as the backbone. Given the considerable noise in low-light images, we compare our results with state-of-the-art denoising algorithms, showcasing the superior performance of our framework.
Breast Cancer is the most common form of cancer in women, majorly occurring in the age group of 40–70 years and the second most common cancer worldwide. There are several advances in imageprocessing techniques ...
详细信息
暂无评论