Visit our application: https://ok1341-image-captioning-app-vx288v.streamlitapp.com
Hello, we are Jonas, Moritz and Ole. Together we want to understand, build and present a caption generator using neural networks. We use this subtype of machine learning because it is closest to the way humans analyze images. How does the application work? The user has the option to either upload an image or select one from the dataset. Our model then tries to create a label that is as accurate as possible. To evaluate, the user can vote via a button which of the two captions they think better represents the image content. The results are displayed in a graph for viewing and further analysis.