Meeting Minutes Recorder

Project Title

Meeting Minutes Recorder

Project Area of Specialization Artificial IntelligenceProject Summary

This is research based project. The project focuses on to record minutes of the meetings. Nowadays this task is
done by the person who is called recorder and their job is to note down meeting minutes and provide the team all of the
information that progress the project. The aim of Meeting Minutes Recorder is to develop a Desktop Application that
has capability to get the task done by saving human’s cost & effort. It will be the process of converting a speech signal
into a sequence of words, removing all of the noises (disturbance) from the input (input will be in the form of speech and the disturbance in the input can be any sort of noise of any object like sound of fan, table, ringtone of the cellphone and other kind of sound), differentiate between the  voices and recognizing the voice of the person. In the end
this application will provide the sequence of text which has the complete data or information that has discussed in the
meeting by all of the members attending the meeting. Information will be shown on screen associated with person’s
name.

Project Objectives

The goal of this project is to reduce human efforts and to record all of the information with the valid prove.
Our project will mainly focus on the following objectives:

1.  It can remove unnecessary noise(s).
2.  It will be able to detect real time speech.
3.  It will differentiate between person’s voices and can recognize them and associates text with person’s name.

Project Implementation Method

The implementation of the project has multiple phases. The flow of the project implementation is:

1. Analyzing the API's and algorithms.
2. Data Generation or Collection.
3. Implement noise removal algorithm.
4. Oversampling the dataset.
5. Find the best model training technique betwwen SVM, NN, CNN, LSTM  and then train model over the dataset. 
6. Implement real time module that could be able to take audio of person as input.
7. Implement sound to text algorithm.
8. Do the prediction using the trained model.
9. Generate file creater module in which audio data can be written down.
10. Design User Interface for this system.

Benefits of the Project

This project can be use in various fileds to write down the information by diffrentiating between the persons. The fileds where it can be use are Journalism, Education, Business etc

Technical Details of Final Deliverable

The technical Deliverables of the project can be classified as:

1. Best feature extraction technique for audio data.
2. Best noise removal algorithm that can only reduce irrelevant audio.
3. Oversample the data, for oversampling find the best nearest neighnbor value.
4. Best model to predict person using the audio.
5. Transfer learning technique will also be used.
6. Written data that has been recorded during the meeting associated with the person's name.

Final Deliverable of the Project Software SystemType of Industry IT Technologies Artificial Intelligence(AI)Sustainable Development Goals Decent Work and Economic Growth, Industry, Innovation and InfrastructureRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	38000
GPU	Equipment	2	15000	30000
Microphone	Miscellaneous	1	8000	8000