Team Agnes presents: EasyDocs - A student friendly Machine learning application like no other. A convenient app for all your study needs: from document/text summarization to text-to-speech to OCR, we've got you covered ;)
In this digital age, when so many of us are experiencing classroom learning via online mode it is absolutely necessary to be able to manipulate and also convert documents from one form to another. Hence Team Agnes presents 'EasyDocs' a student-friendly application that enables users to - summarize documents, recognize text and images and convert text/PDF to audio books. As a fun addition, we have also developed 'EzPz' a chatbot, that can help users navigate and understand the importance of EasyDocs.
Choose from 6 unique components:
-
Summarize hyper-links to various Wikipedia pages, blogs or any page with textual content.
-
Summarize any text document and save yourself the time and effort put into reading your lecture notes or articles.
-
Convert the text in any image into a copy-paste friendly format without spending a dime! Ta-dah welcome to the OCR section.
-
Feed text to EasyDocs and watch it get magically get converted into an audio file that you can play in the click of a button!
-
Upload an entire PDF file and get it read out in your own pace(select 'yes' for slower pace) and download the file to listen to it on the go!
-
Ooof! Too much information about the app right? Worry not! EzPz, the beloved EasyDocs chatbot, comes to the rescue with clear cut instructions as to what each module is used for and how they all work.
EasyDocs has been deployed for commercial usage via Streamlit sharing and the working web application can be found here: https://share.streamlit.io/adityashukzy/easy-docs/main/st_app.py
-
Facing issues with how accurately content/text is captured by OCR and pdf file parser. This leads to certain erroneous words getting included in the text output and audiobook respectively.
-
Loading an audio file for a larger PDF document takes quite a while as the number of minutes increases rapidly depending on the number of pages present in the PDF document.
-
Getting a meticulous summary of the uploaded document/URL isn't plausible as the type of summary generation used in EasyDocs is the extractive kind which gives importance to the most relevant sentences present unlike the abstractive summary which generates the same from scratch and without any reference to the pre-existing document or URL.
|
Aditya Shukla
|
Pooja Ravi
|
MIT © Aditya Shukla
This project is licensed under the MIT License - see the License file for details
Made with ❤️ by Team Agnes