The goal of this project is to experiment with computer vision models. Right now I'm working on building a computer vision model to distinguish between photos of roses in bloom and photos of roses which are not blooming. Currently I have a convolutional neural network model which works at roughly 93% generalization accuracy. I would like to build a transformer model next. Once I have an idea of how each of these models work then my plan is to expand my data set by gathering many more pictures of roses from the iNaturalist data set. I'll then use my model to label these photos and investigate how the flowering time has changed over the last 10 or so years. In particular I'd like to test the hypothesis that the average day that a rose is in bloom has gotten earlier in the year over this time period. I'd also like to generalize the model beyond just roses.