Open krinkere opened 5 years ago
I think its a great thought. We can surely figure out how to combine both these models and produce outstanding results as you just suggested. Once it is implemented in the future, I'll close this issue.
I have gone through the code. please share the steps how to run the code using command line and using GUI
As it works right now, it is either text OR cbir... Do you plan to combine the two models later in the development process? I.e. specify a picture and a short description of what you are looking for. For example, if a user submits a picture of a steering wheel of the car, the application will return all similar round object like a bicycle wheel... Hence specifying "car part" along with the steering wheel would produce much more relevant results.