ImproViz is a cutting-edge application designed to simplify your lectures and enhance productivity. With a user-friendly interface and powerful features, ImproViz helps you manage your learning effectively, by creating diagrams and interactive plots of what is being discussed in class. A visual learners dream, to turn boring lectures into fun illustrations.
- Voice to Text Converstion: Converts voice input to Text from which illustrations can be created. Also Enables people having a hard time following along, a real time subtitle to the lecture going on.
- Real - Time Visualizations: Takes grouped up text and converts it into interactive plots.
- Complex AI Backend: From having a downloaded embedding model, openai gpt calls, to having clustering algorithms our application is a true AI centric application
- Sophisticated UI: Interactive and beautiful UI that not only promotes ease on the eye but also takes care of the needs of the color-blind by implementing a color-blind friendly color pallete. More can be found here - https://davidmathlogic.com/colorblind/#%23D81B60-%231E88E5-%23FFC107-%23004D40
- Accessibility Features: Automatic inscriptions, visual diagrams, and color-blind friendly color pallete for those who have trouble keeping up with the ongoings of a fast-paced classroom.
- Visual Learners: People who have difficulty to handling just auditory information.
- Those Hard of Hearing: People who are not able to keep up with the class due to disabilities.
- The Color-Blind: People who are not able to keep up with the class due to disabilities.
- Those who cannot keep up with accents: People who are not able to keep up with the class due to accent-related issues.
Before installing [Software Name], ensure that you have the following requirements met:
- [Operating System]: Specify the OS versions supported (e.g., Windows 10, macOS 11.0+, Ubuntu 20.04)
- [Hardware]: Any specific hardware needs (e.g., minimum RAM, storage)
- [Dependencies]: Any required packages or software (e.g., Python 3.8, Java JDK)
- OpenAI: https://platform.openai.com/docs/overview
- Assembly AI: https://www.assemblyai.com/?utm_source=google&utm_medium=cpc&utm_campaign=Brand&utm_term=assembly%20ai&gad_source=1&gclid=Cj0KCQjwvpy5BhDTARIsAHSilylkKkkXkGYQ_izyrxhzAn1U8TXkefh3quwPHLLtRnszNp9YkcrAWZkaAgjhEALw_wcB
- Once created the AssemblyAI API Key, in the root directory create a config.js file and paste "const apiKey="{your_api_key}"; with your actual assembly api key
- Download the embedding model from here - https://drive.google.com/file/d/11mcufSNCZbSBguoo_SWLRfAZh2nVVjqV/view
- Unzip the contents
- Rename the folder generated from model.h5 to model_folder
- Bring this folder into the root of your project
- On the first Terminal Run:
python generatevisualizations.py - On the second Terminal Run:
http-server .