photo by the author using Dall-E
- dataset: Kaggle
- Medium corresponding article: Medium link
This is the code for the tutorial on how to build and to deploy a web app with streamlit. Contained here is both the code used for pre-processing the data, training the convolutional neural network model (trained with PyTorch ), building the app, and to deploy on the cloud. Also, here you can find the dataset used.
The app uses an artificial intelligence model to predict the position of yoga in an image. The user can upload an image, press a button, and get the top 5 most likely positions.
The model used is ResNet18, an artificial intelligence model (originally trained on ImageNet) and repurposed for this tutorial.
The app can be found at this link, you can test yourself:
The code is an example of how to develop a computer vision app for classification. Having used transfer learning in this case, the model is easily reusable for other possible tasks.
This project is licensed under the MIT License