Using Label-img, I have successfully labeled my images (dimensions 1100 x 1100 pixels), and am currently training the SSD ResNet50 model (from the TensorFlow 2 Detection Model Zoo). I downloaded the model script (i.e. config file), modified it to include the correct paths to my folders, and am currently training it.

Is this technically considered transfer learning? I have read a lot about it, and I am not quite sure exactly what this is. In the future, I will be learning how to fine-tune these models, but for now, my question is this.

If this is not transfer learning, could anyone provide an example to how it is different from what I am doing?

  • 1
    $\begingroup$ Transfer learning means that you take the weights of pre-trained model and plug them into your model. That is, you have n layers from pre trained model and then m layers of your own model. After this, you train your model. $\endgroup$ – Aray Karjauv Nov 13 '20 at 8:55

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.