Object Detection API: Architecture of Faster-RCNN

[WaqarAliMemon]

trafficstars

Hello everyone,

Could anyone please put a light on the default architecture of Faster-RCNN.

I mean how many Convolutional, pooling and fully connected layers are used in default Object Detection API and how can we change thier number.

Moreover, when we are training the model for something new. Is it still called Pre-trained model?

Thank you in advance.

Best regard,s Waqar Ali Memon

[ojasvisancheti]

I am looking for the same answer anyone got something?