Community
cancel
Showing results for 
Search instead for 
Did you mean: 
Highlighted
Community Manager
268 Views

Compile Deeplab

I’m trying to compile the frozen graph for deeplab from tensorflow models, in particular, the mobilenetv2_coco_voc_trainval variant.

 

I got the input and output from tensorboard and tried compiling with the following command:

 

mvNCCompile frozen_graph.pb -s 12 -in=ImageTensor -on SemanticPredictions -is 513 513 -o deeplab.pb

 

this returned the following error:

 

"

 

shape: (1, ?, ?, 3)

 

Traceback (most recent call last):

 

File "/usr/local/bin/mvNCCompile", line 156, in

 

create_graph(args.network, args.inputnode, args.outputnode, args.outfile, args.nshaves, args.inputsize, args.weights, args.explicit_concat, args.ma2480, args.scheduler, args)

 

File "/usr/local/bin/mvNCCompile", line 137, in create_graph

 

load_ret = load_network(args, parser, myriad_config)

 

File "/usr/local/bin/ncsdk/Controllers/Scheduler.py", line 82, in load_network

 

parse_ret = parse_tensor(arguments, myriad_conf)

 

File "/usr/local/bin/ncsdk/Controllers/TensorFlowParser.py", line 259, in parse_tensor

 

input_data = np.random.uniform(0, 1, shape)

 

File "mtrand.pyx", line 1307, in mtrand.RandomState.uniform

 

File "mtrand.pyx", line 242, in mtrand.cont2_array_sc

 

TypeError: index returned non-int (type NoneType)

 

"

 

Thanks for any help

Tags (1)
0 Kudos
5 Replies
Highlighted
Community Manager
13 Views

I meet the same problem,Do you solve these problems?

0 Kudos
Highlighted
Community Manager
13 Views

@rosenblat Can you provide a link to the model?

0 Kudos
Highlighted
Community Manager
13 Views

0 Kudos
Highlighted
Community Manager
13 Views

@zhanghanbin3159 Unfortunately at the moment, the NCSDK doesn't have support for semantic image segmentation.

0 Kudos
Highlighted
Community Manager
13 Views

@Tome_at_Intel do you have an ETA of NCSDK support for semantic image segmentation?

0 Kudos