Hello all,
I have recently created a fresh fruit classifier app using chapter 2 of part 1. The app works well for banana, apple and orange images. I was not able to deploy my app using Voila and Binder . I have used Streamlit to deploy my app. Works well for me.
App URL:https://share.streamlit.io/rajatpaliwal/fruit_freshness_classifier/main
Github:rajatpaliwal (Rajat Paliwal) · GitHub
Check out the app and provide feedback. Hope you guys have fun!
4 Likes
Hey @RajatP, great work it works really well with real apples and bananas. And it also shows what it thinks about Bananas in Pyjamas 
1 Like
Yeah, they do seem fresh as predicted by the model 
Hm when I use this image it says rottenbanana; Probability: 0.9552 
That’s because I used a slightly different image of pyjamas in bananas
But the probability was very low: Prediction: freshbanana; Probability: 0.4563
I think that’s fair because of the clothes these bananas are wearing.
Oh please don’t get me wrong. I just wanted to fool around and input something funny.
The given image is not showing real fruits, so one can not expect that the model returns a proper prediction.
Here is the link to my first blog https://rajatpaliwal02.medium.com/happiness-and-your-web-app-deployment-e9398974cd71