What I built
I have built an application which can detect whether you are wearing a t-shirt or a shirt via webcam.
Category Submission:
Built for Business
App Link
https://clothdetector-kegak.ondigitalocean.app/
Screenshots
Description
User has to stand in front of the camera wearing a shirt or a t-shirt and then the application will automatically detect the type of clothing apparel. This app asks for camera permission which a user has to allow. All the rendering and detection part is done on client-side. It is basically a client-side application no data is stored on the server.
Link to Source Code
https://github.com/sisodiya2421/ClothDetector
Permissive License
Background
I am currently learning the basics of machine learning and then one day I came across an awesome tool Teachable Machine which lets you build your own custom machine learning model and teaches you the basic concepts of Machine Learning using tensorflow.js. Since I work in a fashion-tech company, from there I got the idea to build something which can detect human clothes using a camera.
How I built it
I built it using node.js and p5js library ( a wrapper around tensorflow.js). DigitalOcean's App Platform gave me a platform to easily deploy and test the performance of the application. Not only this project helped me learn about new libraries like P5js and tensorflow.js but also the javascript language which is new for me. I also learned Key points like epochs and how to attain accuracy during a machine learning training process by building this project.
Latest comments (2)
Well done!
Thank You