[UPDATE]: I will not be looking into issues or emails related to this repo anymore as I am currently very busy with other research commitments. Thank you.
This python application recognizes digits from real time webcam data. The user is allowed to write the digits on the screen using an object-of-interest (a water bottle cap in this case).
The code is in Python (version 3.6 or higher). You also need to install OpenCV and Keras libraries.
A popular demonstration of the capability of deep learning techniques is object recognition in image data.
The "Hello World" of object recognition for machine learning and deep learning is the MNIST dataset for handwritten digit recognition. A set of sample images is shown below.
Each of the digit is stored as a numbered array as shown below.
I built a Multilayer Perceptron (MLP) model as well as a Convolutional Neural Network (CNN) model using Keras library. The predictions of both the models are shown on the screen in real time.
The Test accuracies were as follows:
- MLP Test Accuracy: 98.2%
- CNN Test Accuracy: 99.17%
However, I have observed that the CNN model was doing way better than the MLP when integrated with the webcam data. I do not exactly know why MLP is performing worse than it did with the test data, maybe it has something to do with the resizing I did after grabbing the digit drawn on the screen. I am not sure.
All Hail CNNs !!
Order of Execution is as follows:
Step 1 - Execute python mlp_model_builder.py
Step 2 - Execute python cnn_model_builder.py
Step 3 - This could take a while, so feel free to take a quick nap.
Step 4 - Execute python digits_recognition.py
Step 5 - Grab a blue bottle cap and have fun!