Efficient and low complex architecture for detection and classification of Brain Tumor using RCNN with Two Channel CNN

In today’s world, we find ourselves struggling to fight one of the worst pandemics in the history of humanity known as COVID-2019 caused by a coronavirus. When the virus reaches the lungs, we observe ground-glass opacity in the chest X-ray due to fibrosis in the lungs. Due to the significant differences between X-ray images of an infected and non-infected person, artificial intelligence techniques can be used to identify the presence and severity of the infection. We propose a classification model that can analyze the chest X-rays and help in the accurate diagnosis of COVID-19. Our methodology classifies the chest X-rays into four classes viz. normal, pneumonia, tuberculosis (TB), and COVID-19. Further, the X-rays indicating COVID-19 are classified on a severity-basis into mild, medium, and severe. The deep learning model used for the classification of pneumonia, TB, and normal is VGG-16 with a test accuracy of 95.9 %. For the segregation of normal pneumonia and COVID-19, the DenseNet-161 was used with a test accuracy of 98.9 %, whereas the ResNet-18 worked best for severity classification achieving a test accuracy up to 76 %. Our approach allows mass screening of the people using X-rays as a primary validation for COVID-19.

/pdf/chest-x-ray-classification-using-deep-learning-for-automated-4aezs7dfa7.pdf

Chest X-ray Classification Using Deep Learning for Automated COVID-19 Screening

Identification of malnutrition and prediction of BMI from facial images using real-time image processing and machine learning

Nowadays, the proliferation of automobile industries is directly related to the increasing number of car incidents. So, insurance companies are facing many simultaneous claims and solving claims leakage. The sense of Artificial Intelligence (AI) based on machine learning and deep learning algorithms can help to solve these kinds of problem for insurance industries. In this paper, we apply deep learning-based algorithms, VGG16 and VGG19, for car damage detection and assessment in real-world datasets. The algorithms detect the damaged part of a car and assess its location and then its severity. Initially, we discover the effect of domain-specific pre-trained CNN models, which are trained on an ImageNet dataset, and followed by fine-tuning, because some of the categories can be fine-granular to get our specific tasks. Then we apply transfer learning in pre-trained VGG models and use some techniques to improve the accuracy of our system. We achieve the accuracy of 95.22% of VGG19 and 94.56% of VGG16 in the damaged detection, the accuracy of 76.48% of VGG19 and 74.39% of VGG16 in damage localization, the accuracy of 58.48% of VGG19 and 54.8% of VGG16 in damage severity with the combination of transfer learning and L2 regularization. From their results, the performance of VGG19 is better than VGG16. After analyzing and implementing our models, we find out that the results of using transfer learning and L2 regularization can work better than those of fine-tuning.

Car Damage Detection and Classification

CNL-UNet: A novel lightweight deep learning architecture for multimodal biomedical image segmentation with false output suppression

Traditionally, data mining algorithms and machine learning algorithms are engineered to approach the problems in isolation. These algorithms are employed to train the model in separation on a specific feature space and same distribution. Depending on the business case, a model is trained by applying a machine learning algorithm for a specific task. A widespread assumption in the field of machine learning is that training data and test data must have identical feature spaces with the underlying distribution. On the contrary, in real world this assumption may not hold and thus models need to be rebuilt from the scratch if features and distribution changes. It is an arduous process to collect related training data and rebuild the models. In such cases, Transferring of Knowledge or transfer learning from disparate domains would be desirable. Transfer learning is a method of reusing a pre-trained model knowledge for another task. Transfer learning can be used for classification, regression and clustering problems. This paper uses one of the pre-trained models – VGG 16 with Deep Convolutional Neural Network

Transfer learning using VGG-16 with Deep Convolutional Neural Network for Classifying Images

21st century is touted as a data revolution where tons of data silos being created every second. These data silos are stored across various servers and creating authentic data mines across multiple locations. Business leaders across the world realized the significance of the information residing in these data silos and constantly trying to extract relevant information which can be used to strategize and make savvy decisions. Prodigious salient information out there concealed in customer reviews, tweets, social media comments, and the challenge is to extract hidden insights from it. It is an arduous process to extract meaningful information from these data silos manually. One of the imperative aspects in Natural Language Processing is sentiment Analysis - Extracting emotions, opinions, sentiments, fervor from the corpus of data and showcasing general opinion of the reviewers about the product. All previous works in NLP rely on traditional classification techniques like Support Vector Machines, Naive Bayes, Maximum Entropy, Random forest, etc. With the advent of deep learning models, there is a reliable improvement in accuracy in Natural Language Processing. This paper introduces a deep learning technique namely Deep Convolutional Neural Network which captures a sentiment in the text corpus. The performance of this model will be evaluated on the Amazon product reviews and IMDB movie reviews.

Sentiment Analysis on Customer Reviews using Convolutional Neural Network

Since the last decade, internet plays an imperative and vital role in the creation and retrieval of colossal amounts of information. With ever-increasing advancements in technological field and creation of data at an exponential rate, impertinent or irrelevant data is proliferating at a vast scale in commensuration with relevant data. Moreover, the usage of mobile phones has increased drastically, and phones are becoming an evident part of everyone's lives. With this, there is a notable increase in the number of spam messages from spammers. According to recent statistics, 96% of Indians receive unsolicited text messages every day. SMS spam is any unwanted or unsolicited text note in the form of weblink, promotional message or irrelevant text sent uncritically and non-selectively to your mobile phone, regularly for advertising purposes. The surge in unsolicited information across all platforms including mobile text messages and emails has created an expedited need for the advancement and refinement of more reliable filters to counteract the spam in these messages. Traditionally, rule-based approach is employed to counteract spam messages. According to this approach, a set of rules are employed on the messages by some authority manually. By this method, no favorable or assuring results will be shown because the rules need to regularly be restructured based on the source of spam messages, which is an arduous process. Instead, we use deep learning methods that are efficient and does not require any rules. Deep learning models require a set of training dataset samples to learn the rules from these SMS messages and build a text classifier that efficiently classifies spam from these messages. This paper presents a systematic review of employing deep learning methods namely, convolutional neural network and recurrent neural network on huge corpus of SMS texts to build a spam classifier that classifies messages as ham or spam.

Srikanth Tammina

Papers

Transfer learning using VGG-16 with Deep Convolutional Neural Network for Classifying Images

Sentiment Analysis on Customer Reviews using Convolutional Neural Network

A Comparative Study of Deep Learning Methods for Spam Detection