Computer Vision Project For Determining if an image is a hotdog, or not a hotdog, based on the brilliant App idea from Jian-Yang on "Silicon Valley".
The goal of this project is to train a convolutional neural network model with Keras to reliably distinguish if an image is a hotdog, or not a hotdog. With some luck, it may even be able to answer the age-old question of whether or not a hotdog is a sandwich.
We are attempting to see if the model performs better than blindly guessing (50/50 in binary classification).
Using convolutional neural networks with Max Pooling layers, we have managed to achieve an accuracy of 70% with our model. I have deployed this model on Streamlit to be tested, which takes uploaded images and returns a percentage response for whether or not the model believes it is looking at a hotdog. Test it out and see what results you get!
The images were taken from this dataset on Kaggle.
The idea for this is taken directly from "Silicon Valley". As Jian-Yang, visionary creator of such a profound app, has inspired me, I have decided to use machine learning to recreate the App and see if we can determine if an image is a hotdog, or not a hotdog. Watch "Silicon Valley" on HBO.