images are created in a variety of ways in various industries. These images are tough to work with, and as a result, they can’t be used effectively in a variety of fields. In this paper, image Resolution is improved ...
详细信息
Utilizing photographs and machine vision to accurately determine and categorize fruits throughout growing is critical, not only for reducing employment human morphology data evaluations also for optimizing harvest tas...
详细信息
Any development in technology is made with the intention of solving the difficulties in that field. One such identified problem is blind people were unable to make out the type of currency. The proposed model efficien...
详细信息
In the last few years deep neural networks has significantly improved the state-of-the-art of robotic vision. However, they are mainly trained to recognize only the categories provided in the training set (closed worl...
详细信息
ISBN:
(纸本)9783031064272;9783031064265
In the last few years deep neural networks has significantly improved the state-of-the-art of robotic vision. However, they are mainly trained to recognize only the categories provided in the training set (closed world assumption), being ill equipped to operate in the real world, where new unknown objects may appear over time. In this work, we investigate the open world recognition (OWR) problem that presents two challenges: (i) learn new concepts over time (incremental learning) and (ii) discern between known and unknown categories (open set recognition). Current state-of-the-art OWR methods address incremental learning by employing a knowledge distillation loss. It forces the model to keep the same predictions across training steps, in order to maintain the acquired knowledge. This behaviour may induce the model in mimicking uncertain predictions, preventing it from reaching an optimal representation on the new classes. To overcome this limitation, we propose the Poly loss that penalizes less the changes in the predictions for uncertain samples, while forcing the same output on confident ones. Moreover, we introduce a forget constraint relaxation strategy that allows the model to obtain a better representation of new classes by randomly zeroing the contribution of some old classes from the distillation loss. Finally, while current methods rely on metric learning to detect unknown samples, we propose a new rejection strategy that sidesteps it and directly uses the model classifier to estimate if a sample is known or not. Experiments on three datasets demonstrate that our method outperforms the state of the art.
Eye disease recognition is a challenging task, which usually requires years of medical experience. In this work, we conducted research that can be a start for the most versatile solution. We tried to solve the problem...
详细信息
ISBN:
(纸本)9783031064272;9783031064265
Eye disease recognition is a challenging task, which usually requires years of medical experience. In this work, we conducted research that can be a start for the most versatile solution. We tried to solve the problem of the classification of different eye diseases using neural networks. The firststep of this work consists of gathering all publicly available eye disease datasets and preprocessing them to make the experiments as generalized as possible. This led to the creation of a dataset composed of over 30,000 images. The aim was to teach the model the actual symptoms of the diseases instead of adjusting the results to a given part of the dataset. Several deep convolutional neural networks were used as feature extractors and they were combined with the Synergic Deep learning model. We conducted experiments on the data and were able to achieve promising results.
The proceedings contain 18 papers. The special focus in this conference is on Applied Mathematics and Computational Intelligence. The topics include: steady and Unsteady Solutions of Free Convective Micropolar Fluid F...
ISBN:
(纸本)9789811981937
The proceedings contain 18 papers. The special focus in this conference is on Applied Mathematics and Computational Intelligence. The topics include: steady and Unsteady Solutions of Free Convective Micropolar Fluid Flow Near the Lower stagnation Point of a Solid Sphere;low-Light image Restoration Using Dehazing-Based Inverted Illumination Map Enhancement;mathematical Modeling of Probability and Profit of Single-Zero Roulette to Enhance Understanding of Bets;Nonlinear Computational Crack Analysis of Flexural Deficit Carbon and Glass FRP Wrapped Beams;Economic Benefit Analysis by Integration of Different Comparative Methods for FACTS Devices;optimal Pricing with Servicing Effort in Two Remanufacturing Scenarios of a Closed-Loop Supply Chain;a Review on Type-2 Fuzzy Systems in Robotics and Prospects for Type-3 Fuzzy;Optimization of Palm Oil Mill Effluent (POME) Solubilization Using Linguistic Fuzzy Logic and machinelearning Techniques;analysis of Heat Transfer Coefficients and Pressure Drops in Surface Condenser with Different Baffle Spacings;integration of Big Data and Internet of Things (IoT): Opportunities, Security and Challenges;intuitionistic Fuzzy Metrics and Its Application;complex structure of Number in Language processing;digital Newspaper Using Augmented Reality;nonlocal Fuzzy Solutions for Abstract Second Order Differential Equations;performance Assessment of Routing Protocols in Cognitive Radio Vehicular Ad Hoc Networks;infinite System of Second Order Differential Equations in Banach Space c0.
If overlooked diabetic retinopathy (DR), a degenerative eye condition that affects people with diabetes, can cause serious vision loss. Early DR detection is essential for prompt management and action. Deep convolutio...
详细信息
ISBN:
(纸本)9783031730641
If overlooked diabetic retinopathy (DR), a degenerative eye condition that affects people with diabetes, can cause serious vision loss. Early DR detection is essential for prompt management and action. Deep convolutional neural networks (DCNNs) have become effective tools for automated processing of retinal pictures in recent years, making it possible to diagnose DR earlier. To learn and extract distinguishing features from retinal fundus images, the suggested method makes use of a DCNN architecture. For the training of models and evaluation, a dataset containing quite a few of retinal pictures from diabetic patients—including both normal and DR cases is utilised. Combining methods of supervised learning, the DCNN model undergoes training with a focus on improving performance indicators including accuracy, sensitivity, and specificity. To intentionally enhance the quantity and diversity of the training dataset, methods like image rotation, flipping, scaling and random cropping are used. This successfully simulates changes in the circumstances of collecting images and the clinical traits of DR. The efficiency of the suggested DCNN-based technique for early DR detection is shown by experimental findings. The trained model performs exceptionally well in spotting early indications of DR, such as microaneurysms, haemorrhages, and exudates, with high accuracy. An independent test dataset is used to assess the model’s performance, highlighting its potential for use in practical situations. The suggested approach advances automated DR detection, enabling immediate action and greater control of this illness that threatens eyesight. The suggested DCNN-based approach needs to be further improved and validated, and the resulting system needs to be integrated into the current medical system to facilitate widespread early DR detection and treatment. The model achieved validation accuracy of 76.80% and training accuracy of 99.58%. A training loss of 0.0132 and a validation loss of
machinelearning and artificial intelligence has recently become a prominent technology. Given its popularity and strength in patternrecognition and categorization, many corporations and institutions have begun inves...
详细信息
ISBN:
(纸本)9781607685395
machinelearning and artificial intelligence has recently become a prominent technology. Given its popularity and strength in patternrecognition and categorization, many corporations and institutions have begun investing in healthcare research to improve illness prediction accuracy. Using these strategies, however, has several drawbacks. One of the primary issues is the lack of huge data sets for medical pictures. An introduction to deep learning in medical imageprocessing, from theoretical foundations to real-world applications. The article examines the general appeal of deep learning (DL), a collection of computer science advances. The next step was to learn the basics of neural networks. That explains the use of deep learning and CNNs. So we can see why deep learning is rapidly advancing in various application fields, including medical imageprocessing. The goal of this research was to use innovative methodologies on cancer datasets to explore the feasibility of combining machinelearning and deep learning algorithms for cancer detection. This study used text and picture databases to classify cancer. The datasets are the Liver BUPA disorder database and brain MRI pictures. This article provides optimization methods that outperform the suggested approaches' accuracy. Using two alternative training methods, Levenberg Marquardt (lm) and Resilient back propagation (rp), two classification algorithms were evaluated with different groups of neurons to identify benign and malignant patients. Cascade correlation utilizing the train (rp) outperformed feed forward back propagation using the train (lm). The second deep neural network model presented a technique (based on CNN) for automated brain tumour identification using MRI data. The Water Cycle Algorithm is used to optimise CNN. The established approach is very accurate. The suggested framework examines innovative texture classification algorithms using Discrete Wavelet Transform (DWT) and the Gray Level Co occurrence
The advantages of deepfakes in many applications are counterbalanced by their malicious use, for example, in reply-attacks against a biometric system, identification evasion, and people harassment, when they are wides...
详细信息
ISBN:
(纸本)9783031064302;9783031064296
The advantages of deepfakes in many applications are counterbalanced by their malicious use, for example, in reply-attacks against a biometric system, identification evasion, and people harassment, when they are widespread in social networks and chatting platforms (cyberbullying) as recently documented in newspapers. Due to its "arms-race" nature, deepfake detection systems are often trained on a certain class of deepfakes and showed their limits on never-seen-before classes. In order to shed some light on this problem, we explore the benefits of a multimodal deepfake detection system. We adopted simple fusion rules, which showed their effectiveness in many applications, for example, biometric recognition, to exploit the complementary of different individual classifiers, and derive some possible guidelines for the designer.
As an important topological property of a 3D binary image, the Euler number can be calculated by counting certain 2 x 2 x 2 voxel patterns in the image. This paper presents a novel method for improving the voxel-patte...
详细信息
ISBN:
(纸本)9783031133244;9783031133237
As an important topological property of a 3D binary image, the Euler number can be calculated by counting certain 2 x 2 x 2 voxel patterns in the image. This paper presents a novel method for improving the voxel-pattern-based Euler number computing algorithm of 3D binary images. In the proposed method, by changing the accessing order of voxels in 2 x 2 x 2 voxel patterns and combining the voxel patterns which provide the same Euler number increments for the given image, the average numbers of voxels to be accessed for processing a 2 x 2 x 2 voxel pattern can be decreased from 8 to 4.25, which will lead to an efficient processing. Experimental results demonstrated that the proposed method is much more efficient than the conventional voxel-pattern-based Euler number computing algorithm.
暂无评论