TCS Internship Project
Project Name: Automatic Video Annotation
Automatic annotation of video refers to extraction of the information about video automatically. Here in the project, we approached Deep Learning which is based on Convolutional Neural Networks (CNN) and YOLO Models for Object Recognition rather than approaching Machine learning using Support Vector Machine (SVM). Though CNN accuracy is low comparative to SVM, the time complexity (speed) of CNN is effective which is more preferable for security. For real time computer vision, we enhanced the huge open-source library called Open CV Packages contains more than 2500 optimized algorithms and the language used for coding is Python and Pascal Dataset is used which contains 20 classes. Automatic Video Annotation has been proposed to act as a mediator in the applications of content-based video Retrieval (CBVR) and video management. This study is based on the machine learning technique to annotate video. Hence, the key idea of Automatic Video Annotation (AVA) is to construct the model(s) through automatic learning semantic concept from many videos (even shots or keyframes), then utilize these concept models(s) for predicting appropriate annotation/label for any new video. Later these annotated videos can be retrieves by textual queries.
Academic Projects
Project Name: Optimization of Mini-Max Algorithm with Alpha-Beta Pruning (on Connect4Game)
Evaluation: Kennesaw State University – KSU
Optimization of Connect4 Game using Mini-Max Algorithm and Alpha-Beta Pruning Algorithm. This project evaluated the performance of the algorithm on 3 different metrics likely, Compilation Time Comparison, Space Allocation Comparison, and Accuracy Comparison. In conclusion, this project demonstrates that the alpha-beta pruning algorithm is a better implementation than the mini-max algorithm according to the time complexity scenario.
Project Name: Laser Security Systems
Evaluation: Jawaharlal Nehru Institute of Technology – JNTU
A laser-based Security System is a type of security and alarm system that uses laser light and a light sensor. A security system protects our homes, offices, banks, lockers, etc. from intrusion and unauthorized access
On Interest Projects
Project Name: Image Format Converter
PIL (Python Imaging Library) adds image processing capabilities to python interpreter. The image will be broken into RGB pixels which categorizes the each pixel color definition. However, the color definition of each format of an image changes according to the vector density of that format. Hence, we have to convert the certain formats of an image according to the adjustments of their RGB function. There are different formats of Images for instance, JPEG or JPG, PNG, TIFF, AVIF, APNG, GIF, SVG, BMP, WebP. The PIL Library is capable of processing image according to their RGB ratio. Hence, the source code can change the bulk of image datasets from one format to another format.
Project Name: Natural Language Image Steganography
Steganography is a secret messaging techniques which can be highly implemented in defense or security purpose. Image Steganography is the process of hiding information or data (even code) in an image. As image is the combination of number of pixels having RGB color composition, the written data can be encoded into the pixel formation and can be decoded at the end. The data can be encoded to each pixel with a certain coding format by altering the RGB color composition, the color change can not be noticed though. The future work will be integrated on identifying or detecting whether there is any malicious code in a received image or not.