"You can't Handle the Truth!"

TransferLearning_On_WebBrowser

I made a custom 4-class object classifier using the webcam on the fly. We're going to make a classification through MobileNet model for a particular webcam image and use that for classification.

Built With

• TensorflowJs - TensorFlow.js is a library for developing and training ML models in JavaScript, and deploying in the browser
• Flask - Flask is a micro web framework written in Python.
• MobileNet - MobileNet is an architecture which is more suitable for mobile and embedded based vision applications where there is lack of compute power. This architecture was proposed by Google.

Requirements:

• A recent version of chrome browser.
• A text editor
• Understanding of HTML, CSS, JavaScript and Chrome Dev Tool

Getting Started

• Load index.html page in your Chrome browser.
• you can use common objects or face/body gestures to capture images for each of the three classes(eg: Add A, Add B, Add C).
• Each time you click one of the "Add" buttons, one image is added to that class as a training example. While you do this, the model continues to make predictions on webcam images coming in and shows the results in real-time.
• I have also added a "No Class" button to make sure that No class is predicted other than A, B, C.
• Make Sure that you send single training example by clicking button sequentially from Left to Right.