Building extraction with high accuracy using semantic segmentation from high-resolution remotely sensed imagery has a wide range of applications like urban planning, updating of geospatial database, and disaster management. However, automatic building extraction with non-noisy segmentation map and obtaining accurate boundary information is a big challenge for most of the popular deep learning methods due to the existence of some barriers like cars, vegetation cover and shadow of trees in the high-resolution remote sensing imagery. Thus, we introduce an end-to-end convolutional neural network called Generative Adversarial Network (GAN) in this study to tackle these issues. In the generative model, we utilized SegNet model with Bi-directional Convolutional LSTM (BConvLSTM) to generate the segmentation map from Massachusetts building dataset containing high-resolution aerial imagery. BConvLSTM combines encoded features (containing of more local information) and decoded features (containing of more semantic information) to improve the performance of the model even with the presence of complex backgrounds and barriers. The adversarial training method enforces long-range spatial label vicinity to tackle with the issue of covering building objects with the existing occlusions such as trees, cars and shadows and achieve high-quality building segmentation outcomes under the complex areas. The quantitative results obtained by the proposed technique with an average F1-score of 96.81% show that the suggested approach could achieve better results through detecting and adjusting the difference between the segmentation model output and the reference map compared to other state-of-the-art approaches such as autoencoder method with 91.36%, SegNet+BConvLSTM with 95.96%, FCN-CRFs with 95.36%% SegNet with 94.77%, and GAN-SCA model with 96.36% accuracy.

No abstract

The stock market is very complex and volatile. It is impacted by positive and negative sentiments which are based on media releases. The scope of the stock price analysis relies upon ability to recognise the stock movements. It is based on technical fundamentals and understanding the hidden trends which the market follows. Stock price prediction has consistently been an extremely dynamic field of exploration and research work. However, arriving at the ideal degree of precision is still an enticing challenge. In this paper, we are proposing a combined effort of using efficient machine learning techniques coupled with a deep learning technique—Long Short Term Memory (LSTM)—to use them to predict the stock prices with a high level of accuracy. Sentiments derived by users from news headlines have a tremendous effect on the buying and selling patterns of the traders as they easily get influenced by what they read. Hence, fusing one more dimension of sentiments along with technical analysis should improve the prediction accuracy. LSTM networks have proved to be a very useful tool to learn and predict temporal data having long term dependencies. In our work, the LSTM model uses historical stock data along with sentiments from news items to create a better predictive model.

Intelligent transportation systems (ITSs) are one of the most widely-discussed and researched topic across the world. The researchers have focused on the early prediction of a driver's movements before drivers actually perform actions, which might suggest a driver to take a corrective action while driving and thus, avoid the risk of an accident. This article presents an improved deep-learning technique to predict a driver's action before he performs that action, a few seconds in advance. This is considering both the inside context (of the driver) and the outside context (of the road), and fuses them together to anticipate the actions. To predict the driver's action accurately, the proposed work is inspired by recent developments in recurrent neural networks (RNN) with long short term memory (LSTM) algorithms. The performance merit of the proposed algorithm is compared with four other algorithms and the results suggest that the proposed algorithm outperforms the other algorithms using a range of performance metrics.

Background Any contamination in the human body can prompt changes in blood cell morphology and various parameters of cells. The minuscule images of blood cells are examined for recognizing the contamination inside the body with an expectation of maladies and variations from the norm. Appropriate segmentation of these cells makes the detection of a disease progressively exact and vigorous. Microscopic blood cell analysis is a critical activity in the pathological analysis. It highlights the investigation of appropriate malady after exact location followed by an order of abnormalities, which assumes an essential job in the analysis of various disorders, treatment arranging, and assessment of results of treatment. Methodology A survey of different areas where microscopic imaging of blood cells is used for disease detection is done in this paper. Research papers from this area are obtained from a popular search engine, Google Scholar. The articles are searched considering the basics of blood such as its composition followed by staining of blood, that is most important and mandatory before microscopic analysis. Different methods for classification, segmentation of blood cells are reviewed. Microscopic analysis using image processing, computer vision and machine learning are the main focus of the analysis and the review here. Methodologies employed by different researchers for blood cells analysis in terms of these mentioned algorithms is the key point of review considered in the study. Results Different methodologies used for microscopic analysis of blood cells are analyzed and are compared according to different performance measures. From the extensive review the conclusion is made. Conclusion There are different machine learning and deep learning algorithms employed by researchers for segmentation of blood cell components and disease detection considering microscopic analysis. There is a scope of improvement in terms of different performance evaluation parameters. Different bio-inspired optimization algorithms can be used for improvement. Explainable AI can analyze the features of AI implemented system and will make the system more trusted and commercially suitable.