Delhi Technological University

About: Delhi Technological University is a education organization based out in New Delhi, India. It is known for research contribution in the topics: Computer science & Control theory. The organization has 4427 authors who have published 6761 publications receiving 71035 citations. The organization is also known as: Delhi College of Engineering & DTU.

Disease Detection in Coffee Plants Using Convolutional Neural Network

Abstract: Every year, the crop output of India is significantly affected due to delayed detection of diseases in crops. This research is a contribution to the farmers in their battle against coffee plants diseases. It will help in timely detection of diseases, resulting in increased coffee production output of India. Many coffee plant diseases like Leaf Rust, Cercospora Spots have clear visual symptoms and thus can be extracted and their classification can be done. Convolutional Neural Networks (CNNs) has proved its efficiency and accuracy in the field of image classification and pattern recognition. Hence it can act as a powerful tool in the diagnosis of coffee leaves diseases since these symptoms have clearly distinguishable patterns. Thus, a Convolution Neural Network model is proposed which utilizes the technique of Transfer Learning, reducing the training time taken by the model significantly. Further, to achieve a higher success rate, Data Augmentation technique is applied to enlarge the dataset used to train the network. The proposed model has achieved a high accuracy of 97.61%.

Can luteolin be a therapeutic molecule for both colon cancer and diabetes

Abstract: Diabetes and colon cancer are the leading cause of mortality worldwide. According to World Health Organization, the number of patients with diabetes and cancer is going to be elevated by 50% in 2020. However, several flavonoids have been known to be useful in reducing the chance of cancer/diabetes but the hunt of a single biomolecule that can act as therapeutic and preventive molecules for future epidemic continues. In this review, we aim to perform an illustration of all researches done that target molecular signaling using luteolin in cancer/diabetes and predicted target protein using PharmMapper. The search confirms that luteolin can be a remedial molecule for both cancer and diabetes via acting on variety of signaling pathway. Furthermore, we also intend to illustrate/compare the predicted and verified molecular modes of action of luteolin. Fluorescence in situ hybridization analysis confirms the expression of CCND1 in colon cancer while immunofluorescence analysis confirms the CDK4 in diabetes. Finally, an effort has been made to map docking of marker protein-luteolin at a particular site using docking software. This review gives a holistic overview about luteolin as a therapeutic molecule for cancer/diabetes via acting on multiple signaling cascade such as p53, Wnt, eNOS, iNOS, SOD and MMP9, with especial emphasis on the cyclin-CDK pathway. Altogether, the review concludes that luteolin can be a molecule for the therapy of both cancer and diabetes by acting on broad signaling pathway.

Unified framework for human activity recognition: An approach using spatial edge distribution and ℜ-transform

Abstract: In this paper, a unified approach for the recognition of human activity using the spatial edge distribution of gradients and orientation of the human silhouettes in a video sequence is presented. The spatial edge distribution is computed on still image at different levels of resolution of sub-images to extract out the shape of the activity posture. The fuzzy trapezoidal membership function is used to extract the key frames of the activity, and the single still key image is extracted according to the histogram distance. The temporal content of the activity is extracted by the computation of orientation of the silhouettes using ℜ-transform. The ℜ-transform is applied on the binary human silhouettes, and the extraction of human silhouettes from the video sequence is done using texture based segmentation techniques. The high dimensionality of the ℜ-transform features is handled by applying Local linear embedding (LLE) dimension reduction approach. A unified model is constructed by integrating the spatial edge distribution of gradients and temporal content of the activity. The performance of the developed model is demonstrated on publicly available datasets, and the highest classification accuracy achieved on each datasets is compared with the similar state-of-the-art techniques and shows the superior performance.