Seeking internships to apply diverse skills and knowledge gained through in Research, Master's in Computer Science and a Bachelor's in Computer Science. I aim to contribute effectively to projects at the intersection of technology and engineering, fostering innovation and optimizing operations. Passionate towards research, programming, problem solving and building solutions to business problems.
undertook a comprehensive research project focusing on enhancing abnormal activity detection in offline surveillance footage. The culmination of this effort resulted in a published research paper in the esteemed International Journal for Innovative Engineering and Management Research (Volume-12).I developed this using machine and deep learning algorithms [>95%] accuracy which increased my experience in challenging technologies. Published research paper in a reputable international journal, underscoring the significance and quality of my research.
SELFIE CAPTURING BY DETECTING SMILE, We can auto capture selfie by detecting smile. In addition to that feature we add photo capture, browsing option, filters, video option, watermark on image features in this project to make the interface user-friendly. The main idea is to develop and interface similar to the mobile camera in laptop., Python Programming Language (OpenCV library), Atom (IDE)
SMART HIRING WEB APPLICATION(ReactJS), It is a job hiring portal developed using ReactJS, where we can build our profile and apply to various job roles., ReactJS, VS CODE(editor)
REAL AND FAKE FACE DETECTION, The model enables us to detect whether the given input image of a person is a real face or a fake face. The model is trained by more than 10,000 dataset images., Convolutional Neural Network(CNN is a network architecture for deep learning) [99%] accuracy.
ANALYSIS OF ABNORMAL ACTIVITY DETECTION IN OFFLINE SURVEILLANCE FOOTAGE, We are using three separate models in the method to identify the videos of anomaly. The models are convolutional neural networks (CNN), architecture VGG16 and ResNet50. In this project, we use videos of real-time anomalies to train the pattern. So that we can able to evaluate each model's performance independently and use the model to predict anomalies with the highest degree of accuracy. VGG16[98%] model is chosen from the three models for the prediction based on its performance in training and testing., 2D-convolutional neural networks (CNN)[99%], deep architectures of convolutional neural networks VGG16 and ResNet50.