I am using tensorflow for feature extraction from images. Images are divided into 5 categories, each having various designs (which are labelled and trained to the system).

Model is trained to find similar images from the database, when one image is uploaded. For this, I have converted the entire database into 5 vectors (1 for each category).

Everything is working good till here. However, I receive 10000 new images daily which I want to be added to the database for comparison. According to my research, the Vector File that has been created earlier is some kind of chronological file, where, I can't add new image vectors. That is, I will have to re-vectorize the entire Database + 10000 images daily, to revise and receive new Vector File (which is used for comparison)

Question: Is there a way to add new image files (vectors) to this existing vector file which has all the image vectors of the database?

  • $\begingroup$ How are you extracting feature using TF? $\endgroup$
    – 10xAI
    Apr 28 at 15:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.