Training tensorflow object detection api, is it possible to use already cut out photos with an object instead of selecting an object in a photo for training?

V

V

viktor40182020-12-04 15:02:15

Neural networks

viktor4018, 2020-12-04 15:02:15

Training tensorflow object detection api, is it possible to use already cut out photos with an object instead of selecting an object in a photo for training?

At the moment, in order to train MaskRcnn, I have to select an object in the image using programs like labelme or the VGG Image Annotation Tool, but I have sliced pictures with the object I need. Can I somehow use them when teaching or is the background also necessary for teaching?

Reply

Answer the question

Answer the question

In order to leave comments, you need to log in

0 answer(s)

Similar questions

L

lexstile2019-11-22 15:00:47

How much data is needed to train a neural network? 2Reply

B

beduin012015-07-18 20:34:11

Neural network training time? 3Reply

D

di15186112021-05-18 01:56:43

How to create pictures in bitmap binary (-1,1) format? 0Reply

F

frontjss2019-12-03 19:29:14

What is weight in neural networks? 2Reply

X

Xvir432019-12-07 23:22:52

How to drive input into pfan_type? 0Reply

Didn't find what you were looking for?

Ask your question

Ask a Question

731 491 924 answers to any question