Not everyone is going to build a image recognition model from scratch for small scale implementations. And we aren’t sure to start off because there are many image recognition frameworks available on Github. Here is one of the stable framework and also is open to contributions.
Kur Framework for Image Recognition -> https://github.com/deepgram/kur
Good part about Kur is everything is configured through a single yaml file. Knowledge of basic CNN and Pickle is enough to build a good model. It is also easy to setup the TRAIN-TEST-VALIDATE workflow.
Share your thoughts, if there are some other good resources on image recognition.