Inspiration
- Want to educate ppl on Dinos and also harness the power of dnn model
- Want to custom train model and have process of fine tuning it
What it does
- Let users upload dino images and get the predicted species of dino as well as information about its diet, location, period, fun fact, etc.
How we built it
Used python for image classifier model
- Added several different layers with various activation functions and units for the baseline model.
- Went through hyper parameter turning to find the best parameter and came up with the final model ## Challenges we ran into
- Processing our images
- Improving the accuracy of the model ## Accomplishments that we're proud of
- Deploying the model to web app
- Custom training the DNN model ## What's next for Do you no your Dinos? -- Dinosaur DNN Image Classifier
- Need more image preprocessing
- Need further hyper parameter testing
- Need Larger dataset and more time
We couldn't upload the saved model (the one we used to deploy on the web app) to the github and this submission since the size was too big (over threshold for github).
Built With
- flask
- html/css
- keras
- matplotlib
- numpy
- pandas
- python
- scikit-learn
- tensorflow
Log in or sign up for Devpost to join the conversation.