search

bendangnuksung / mrcnn_serving_ready

🛠 Mask R-CNN Keras to Tensorflow and TFX models + Serving models using TFX GRPC & RESTAPI
https://github.com/matterport/Mask_RCNN
MIT License
103 stars 35 forks source link

How can I read input and output node from frozen inference model (.pb)? #25

Open K-M-Ibrahim-Khalilullah opened 4 years ago

K-M-Ibrahim-Khalilullah commented 4 years ago

@bendangnuksung Thanks for your great repository. it helps me lot.

Would you suggest me please How can I read input and output node/tensor from frozen inference graph (.pb)? Thanks again

bendangnuksung commented 4 years ago

Hi @ibrahimLearning How is it different from Issues #9? Judging from your comment on #9 you seem to have successfully able to use the frozen model.

K-M-Ibrahim-Khalilullah commented 4 years ago

Hi @ibrahimLearning How is it different from Issues #9? Judging from your comment on #9 you seem to have successfully able to use the frozen model.

@bendangnuksung Thanks for your reply Yes I used the frozen model using forward.py, what you provided. It was used for detection. In the internal code, the following segment mentions output nodes: self.outputs = { 'detection': 'mrcnn_detection_4/Reshape_1', 'class': 'mrcnn_class_4/Reshape_1', 'box': 'mrcnn_bbox_4/Reshape', 'mask': 'mrcnn_mask_4/Reshape_1'}

But my problem is to display or read the node input and output from .pb file, where I don't know before the name of the model. Is there any clue please?

Thanks