OCR – Document Digitization
Using Computer Vision, we have developed a document digitization tool based on Optical Character Recognition techniques which can easily solve the problem by automating the process of extracting the text out of a digital image or a scanned copy. This can help any organization to keep records more securely and organized.
Using NLP techniques and transformers, we have designed the text summarizer. It is the technique of shortening long pieces of text with the goal to create a coherent and fluent summary having only the main points outlined in the document. There are important applications for text summarization in various NLP related tasks such as text classification, Q&A, legal/medical texts summarization, and news summarization, etc.
Using a semi-supervised approach and active learning, we have developed this tool to minimize the time and effort it takes to overcome the pitfall of labeling data, manually. With the use of this tool, a user can label few data points, and based on those the rest of the data will be annotated. This tool can reduce the development time for a lot of data science projects and make the process much easier.
Brain Cancer Prognosis
Using different image processing techniques and a modified CNN architecture that can predict the methylation status based on the MRI of the Glioma patients, we have developed this application. The usual procedure involves a biopsy which is an invasive as well as a risky process. This model basically bypasses the need for a risky procedure like biopsy and produces results as par with the human diagnosis.