Voice Diary: Emotion Tracking with Speech Recognition

Voice Diary is a diary application that converts user speech into text and analyzes emotions to record them. It visually displays emotions and helps users track their emotional patterns over time 😃

Table of Contents

1. Features
2. Technology Stack
3. Installation
   • Setup Environment
4. Execution
   • Demo
   • User Interface Overview
   • How It Works

Features

• Voice Recording: Record user speech and send it to the server.
• Text Conversion: Convert speech to text.
• Emotion Analysis: Analyze emotions from the converted text.
• Calendar View: Visualize emotional data in a calendar format.
• History Tracking: View text and emotions for specific dates.

Technology Stack

• Backend: Flask (Python)
• Frontend: HTML, CSS, JavaScript
• Database: JSON file for data storage
• Cloud Services:
  • Google Cloud Speech-to-Text API
  • Google Cloud Natural Language API

Installation

Setup Environment

1. Install the required packages

   pip install flask google-cloud-speech google-cloud-language pydub

2. Install FFmpeg

• Download and install FFmpeg, then add it to your system PATH.

3. Set up Google Cloud Service Account

• Generate a service account key (JSON) from Google Cloud Console

Execution

Demo

demo.mp4

User Interface Overview

1. Main Page: The homepage where users can record their voice, analyze emotions, and navigate to past records or the calendar.
2. Past Record: A page showing previous transcriptions and their corresponding emotions for specific dates.
3. Emotion Calendar: A calendar view displaying the emotions recorded for each day.

How It Works

1. Recording

   • Click the "Recording" button to start recording your voice.
   • Click "Stop Recording" when you're done.
   • The recorded audio is sent to the server for transcription and emotion analysis.

2. Text Conversion

   • The recorded voice is converted into text using Google Cloud Speech-to-Text API.

3. Emotion Analysis

   • The transcription is analyzed using Google Cloud Natural Language API to determine the emotion (happy 😄, sad 😢, angry 😡, neutral😐).

4. Data Visualization

   • The transcription and emotion are saved with the current date.
   • Users can view emotions in a calendar view or check detailed past records.