https://github.com/kartikmehta8/textify
Textify is a Speech-To-Text Website that can be operated by gestures of your hands for Starting, Stopping & Aborting the Speech Operations.
https://github.com/kartikmehta8/textify
javascript speech-to-text tensorflow
Last synced: about 2 months ago
JSON representation
Textify is a Speech-To-Text Website that can be operated by gestures of your hands for Starting, Stopping & Aborting the Speech Operations.
- Host: GitHub
- URL: https://github.com/kartikmehta8/textify
- Owner: kartikmehta8
- Created: 2022-03-06T03:31:28.000Z (over 4 years ago)
- Default Branch: main
- Last Pushed: 2022-03-06T07:51:57.000Z (over 4 years ago)
- Last Synced: 2025-01-16T05:55:19.204Z (over 1 year ago)
- Topics: javascript, speech-to-text, tensorflow
- Language: HTML
- Homepage:
- Size: 50.8 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Textify - Gesture Controlled Speech-to-Text Website
Textify is a Speech-To-Text Website that can be operated by gestures of your hands for Starting, Stopping & Aborting the Speech Operations.
## What it is capable of doing?
You can convert **Speech-to-Text or vice versa**. The website can perform all the functions by only governing it with various gestures. Users can Start, Stop or command the website to convert the provided text into speech. The only requirement is that you have to provide access to the camera & microphone. The rest is magic!
## How do I build it?
### The website is built using:
- HTML
- CSS
- Tailwind CSS
- Javascript
- Tensorflow.js
- Google's Teachable Machine
- Google Cloud
- Google Chrome
A user can click **Start with Gestures** to enable the voice recording. There are three gestures for starting, stopping, and converting text-to-speech data. Users can also manually start/stop/play the commands using the buttons provided.
## The challenges I ran into:
- The major challenge was the integration of Google's Teachable Machine with our camera for Gesture Recognition. Thanks to the document that was on the official website.
- Working with webkitSpeechRecognition() was quite uncomfortable initially.
- Setting up coordination between the gestures and speech commands was time-consuming.
## My Learnings & Accomplishments:
1. Convert Speech to Text or vice versa with ease.
2. I integrated an Object Detection Machine Learning Model on a website.
3. Control a website with gestures.
4. I learned how to host the ML Model on Google Cloud using Google's Teachable Machine.
5. Worked with webkitSpeechRecognition().
## What's next for Textify - Gesture Controlled Speech-to-Text Website?
- I will improve the website by deploying it on Google Virtual Machines and making my ML Model self-train as the number of users increases. It will improve the accuracy of the gesture's identification.
- I will also try to create my voice assistant rather than using webkitSpeechRecognition() as it depends from browser to browser.