Data Augmentation Technology Driven By Image Style Transfer in Self-Driving Car Based on End-to-End Learning
8
Citation
44
Reference
10
Related Paper
Citation Trend
Abstract:
With the advent of deep learning, self-driving schemes based on deep learning are becoming more and more popular. Robust perception-action models should learn from data with different scenarios and real behaviors, while current end-to-end model learning is generally limited to training of massive data, innovation of deep network architecture, and learning in-situ model in a simulation environment. Therefore, we introduce a new image style transfer method into data augmentation, and improve the diversity of limited data by changing the texture, contrast ratio and color of the image, and then it is extended to the scenarios that the model has been unobserved before. Inspired by rapid style transfer and artistic style neural algorithms, we propose an arbitrary style generation network architecture, including style transfer network, style learning network, style loss network and multivariate Gaussian distribution function. The style embedding vector is randomly sampled from the multivariate Gaussian distribution and linearly interpolated with the embedded vector predicted by the input image on the style learning network, which provides a set of normalization constants for the style transfer network, and finally realizes the diversity of the image style. In order to verify the effectiveness of the method, image classification and simulation experiments were performed separately. Finally, we built a small-sized smart car experiment platform, and apply the data augmentation technology based on image style transfer drive to the experiment of automatic driving for the first time. The experimental results show that: (1) The proposed scheme can improve the prediction accuracy of the end-to-end model and reduce the model’s error accumulation; (2) the method based on image style transfer provides a new scheme for data augmentation technology, and also provides a solution for the high cost that many deep models rely heavily on a large number of label data.Keywords:
Normalization
Transfer of learning
Lung Cancer is the deadliest type of cancer. For diagnosing and detecting lung cancer, many Deep learning techniques are proposed. Deep learning techniques depend on massive data and extracting 3D features of images, which is a tedious task. Therefore deep transfer learning is used as an alternative for better results in medical image analyses like lung cancer, where pre-trained deep learning models are employed, which resolves the problem of labeled data scarcity. As no systematic survey is available in the literature, this paper reviews Lung Cancer detection and diagnosis using Deep Transfer Learning techniques. Moreover, all such methods are compared based on performance parameters like accuracy, Recall, F1- score and precision, etc. The comparison shows that the overall performance of deep learning techniques with the use of Transfer learning is enhanced.
Transfer of learning
Cancer Detection
Cite
Citations (6)
Purpose The paper aims to address the tracking algorithm based on deep learning and four deep learning tracking models developed. They compared with each other to prevent collision and to obtain target tracking in autonomous aircraft. Design/methodology/approach First, to follow the visual target, the detection methods were used and then the tracking methods were examined. Here, four models (deep convolutional neural networks (DCNN), deep convolutional neural networks with fine-tuning (DCNNFN), transfer learning with deep convolutional neural network (TLDCNN) and fine-tuning deep convolutional neural network with transfer learning (FNDCNNTL)) were developed. Findings The training time of DCNN took 9 min 33 s, while the accuracy percentage was calculated as 84%. In DCNNFN, the training time of the network was calculated as 4 min 26 s and the accuracy percentage was 91%. The training of TLDCNN) took 34 min and 49 s and the accuracy percentage was calculated as 95%. With FNDCNNTL, the training time of the network was calculated as 34 min 33 s and the accuracy percentage was nearly 100%. Originality/value Compared to the results in the literature ranging from 89.4% to 95.6%, using FNDCNNTL, better results were found in the paper.
Transfer of learning
Tracking (education)
Cite
Citations (16)
When using deep learning for DDoS attack detection, there is a general degradation in detection performance due to small sample size. This paper proposes a small-sample DDoS attack detection method based on deep transfer learning. First, deep learning techniques are used to train several neural networks that can be used for transfer in DDoS attacks with sufficient samples. Then we design a transferability metric to compare the transfer performance of different networks. With this metric, the network with the best transfer performance can be selected among the four networks. Then for a small sample of DDoS attacks, this paper demonstrates that the deep learning detection technique brings deterioration in performance, with the detection performance dropping from 99.28% to 67%. Finally, we end up with a 20.8% improvement in detection performance by deep transfer of the 8LANN network in the target domain. The experiment shows that the detection method based on deep transfer learning proposed in this paper can well improve the performance deterioration of deep learning techniques for small sample DDoS attack detection.
Transfer of learning
Sample (material)
Transferability
Cite
Citations (17)
The fast spread of Coronavirus (COVID-19) has created a global health crisis. The World Health organization (WHO) released several recommendations to help in preventing the spread of coronavirus. Wearing a mask in crowded venues is the most appropriate protective practice against COVID-19, according to the WHO. Keeping a close eye on people in public places is next to impossible, so identifying face masks becomes critical in the fight against COVID-19. Medical image analysis and classification are two areas where deep learning has lately become one of the widely used ways for enhancing performance. It can be utilized quite efficaciously to identify individuals who are not wearing a mask. The utilization of transfer learning models as a deep learning technique is also on the rise, and they work quite well. In our study, we used a total number of 7235 data images from an online dataset for face mask detection, which is done with two deep learning models and one transfer learning model. The deep learning-based CNN model using MaxPooling operation and AveragePooling operation achieved the accuracy of 95.78% and 95.36% respectively. Contradictory to that, the transfer learning-based MobileNetV2 model achieved an accuracy of 99.10%.
Transfer of learning
Face masks
Cite
Citations (2)
Training data sparsity is a common problem for many real-world applications in Search and Recommendation domains. Even for applications with a lot of training data, in the cold-start scenario we usually do not get enough labeled data. Transfer Learning is a promising approach for addressing this problem. In addition, features might interact with each other in a complex way that traditional approaches might not be able to represent, Deep Transfer Learning, which leverages Deep Neural Networks for Transfer Learning, might be able to catch such deep patterns hidden in complex feature interactions. Due to these reasons, recently Deep Transfer Learning research has gained a lot of attention and has been successfully applied to many real-world applications. This tutorial offers an overview of Deep Transfer Learning approaches in Search and Recommendation domains from the industry perspective. In this tutorial We first introduce the basic concepts and major categories of Deep Transfer Learning. Then we focus on recent developments of Deep Transfer Learning approaches in the Search and Recommendation domains. After that we will introduce two real-world examples of how to apply Deep Transfer Learning methods to improve Search and Recommendation performance at LinkedIn. Finally we will conclude the tutorial with discussion of future directions.
Transfer of learning
Deep Neural Networks
Feature (linguistics)
Cite
Citations (2)
With the rapid development technology, Artificial Intelligence is the most powerful technique, it has made great progress in many areas, including computer vision and medical imaging. This paper proposes a deep learning-based framework for COVID-19 detection. Deep transfer learning models-based on a pre-trained Deep convolutional Neural Network are proposed. Several pre-trained models, such as DensNet201, InceptionV3, VGG16, and ResNet50 were evaluated for this analysis.The datasets used in this paper for training model are a mix of X-ray and CT images in two distinct categories: Normal and COVID-19. The experimental results proved that the DensNet201 was the most suitable deep transfer model according to the test accuracy measure and that it reached 97% with the other performance metrics such as F1 score, precision, and recall.
Transfer of learning
Cite
Citations (9)
Deep learning has made spectacular achievements in analysing natural images, but it faces challenges for medical applications partly due to inadequate images.Aiming to classify malignant and benign pulmonary nodules using CT images, we explore different strategies to utilize the state-of-the-art deep convolutional neural networks (CNN).Experiments are conducted using the Lung Image Database Consortium image collection (LIDC-IDRI), which is a public database containing 1018 cases. Three strategies are implemented including to 1) modify some state-of-the-art CNN architectures, 2) integrate different CNNs and 3) adopt transfer learning. Totally, 11 deep CNN models are compared using the same dataset.Study demonstrates that, for the model modification scheme, a concise CifarNet performs better than the other modified CNNs with more complex architectures, achieving an area under ROC curve of AUC = 0.90. Integrated CNN models do not significantly improve the classification performance, but the model complexity is reduced. Transfer learning outperforms the other two schemes and ResNet with fine-tuning leads to the best performance with an AUC = 0.94, as well as the sensitivity of 91% and an overall accuracy of 88%.Model modification, model integration, and transfer learning can play important roles to identify and generate optimal deep CNN models in classifying pulmonary nodules based on CT images efficiently. Transfer learning is preferred when applying deep learning to medical imaging applications.
Transfer of learning
Cite
Citations (54)
Abstract COVID -19, is a deadly, dangerous and contagious disease caused by the novel corona virus. It is very important to detect COVID-19 infection accurately as quickly as possible to avoid the spreading. Deep learning methods can significantly improve the efficiency and accuracy of reading Chest X-Rays (CXRs). The existing Deep learning models with further fine tune provide cost effective, rapid, and better classification results. This paper tries to deploy well studied AI tools with modification on X-ray images to classify COVID 19. This research performs five experiments to classify COVID-19 CXRs from Normal and Viral Pneumonia CXRs using Convolutional Neural Networks (CNN). Four experiments were performed on state-of-the-art pre-trained models using transfer learning and one experiment was performed using a CNN designed from scratch. Dataset used for the experiments consists of chest X-Ray images from the Kaggle dataset and other publicly accessible sources. The data was split into three parts while 90% retained for training the models, 5% each was used in validation and testing of the constructed models. The four transfer learning models used were Inception, Xception, ResNet, and VGG19, that resulted in the test accuracies of 93.07%, 94.8%, 67.5%, and 91.1% respectively and our CNN model resulted in 94.6%.
Transfer of learning
Deep Neural Networks
Cite
Citations (2)
Deep Learning has advanced to a greater level in the field of Artificial Intelligence in recent years, and it is currently employed globally. This aids the system in improving its accuracy. Deep Learning algorithms have made image classification considerably more viable, allowing us to analyse large datasets. Deep Convolutional Neural Networks are used in the majority of image classification nowadays. In this paper, Image Classification is performed using the VGG16, ResNet18, ResNet50, GoogleNet, MobileNet, AlexNet in Best Artworks of All Time Dataset which is taken from the Kaggle and the best model for training the dataset is choosen. This Dataset is the collection of the 8355 high resolution portraits which is in form of the RGB Images. After experimentation it is found that, in the Best Artworks of all Time data the ResNet50 achieved better accuracy of 87.15% and loss of 0.0015% among all other trained Deep Networks.
Transfer of learning
Contextual image classification
RGB color model
Cite
Citations (4)
In the current COVID-19 pandemic situation, there is an urgent need to screen infected patients quickly and accurately. Using deep learning models trained on chest X-ray images can become an efficient method for screening COVID-19 patients in these situations. Deep learning approaches are already widely used in the medical community. However, they require a large amount of data to be accurate. The open-source community collectively has made efforts to collect and annotate the data, but it is not enough to train an accurate deep learning model. Few-shot learning is a sub-field of machine learning that aims to learn the objective with less amount of data. In this work, we have experimented with well-known solutions for data scarcity in deep learning to detect COVID-19. These include data augmentation, transfer learning, and few-shot learning, and unsupervised learning. We have also proposed a custom few-shot learning approach to detect COVID-19 using siamese networks. Our experimental results showcased that we can implement an efficient and accurate deep learning model for COVID-19 detection by adopting the few-shot learning approaches even with less amount of data. Using our proposed approach we were able to achieve 96.4% accuracy an improvement from 83% using baseline models.
Transfer of learning
Baseline (sea)
Cite
Citations (1)