Forums - snpe-tensorflow-to-dlc ssd_mobilenet_v1 ValueError: No op named NonMaxSuppressionV3 in defined operations

1 post / 0 new
snpe-tensorflow-to-dlc ssd_mobilenet_v1 ValueError: No op named NonMaxSuppressionV3 in defined operations
tmeharizghi
Join Date: 12 Mar 18
Posts: 4
Posted: Wed, 2018-08-15 16:42

Hi,

Summary:

When converting ssd_mobilenet_v1 model to dlc, I got this error "ValueError: No op named NonMaxSuppressionV3 in defined operations"

Description:

I have downloaded a ssd_mobilenet_v1 pretrained model based on coco dataset from http://download.tensorflow.org/models/object_detection/ssd_mobilenet_v1_coco_2018_01_28.tar.gz. Then I created my own dataset and used transfer learning to retrain with my own dataset of 5 classes. I exported and froze the retrained graph. All these were done on a server with:

OS: Ubuntu 16.04

GPU/CPU: GPU

Tensorflow Version: 1.7.0

Finally I tried to convert the frozen graph to `dlc` on a local computer with:

SNPE Version: 1.18.0.

Tensorflow: 1.6.0

GPU/CPU: CPU

Python: 2.7

I used this command to convert the frozen model:

snpe-tensorflow-to-dlc --graph frozen_inference_graph.pb -i Preprocessor/sub 1,300,300,3 --out_node detection_classes --out_node detection_boxes --out_node detection_scores --dlc ssd_mobilenet_v1.dlc --allow_unconsumed_nodes

However after starting the process, I got the following error:

2018-08-15 16:31:44,221 - 109 - ERROR - Encountered Error: No op named NonMaxSuppressionV3 in defined operations.
Traceback (most recent call last):                                                 
File "/home/tes/snpe-sdk/bin/x86_64-linux-clang/snpe-tensorflow-to-dlc", line 99,
in main                                                                           
model = loader.load(args.graph, in_nodes, in_dims, args.in_type, args.out_node,
session)                                                                          
File "/home/tes/snpe-sdk/lib/python/converters/tensorflow/loader.py", line 50, in
load                                                                              
graph_def = self.__import_graph(graph_pb_or_meta_path, session, out_node_names)
File "/home/tes/snpe-sdk/lib/python/converters/tensorflow/loader.py", line 108, i
n __import_graph                                                                   
tf.import_graph_def(graph_def, name="")                                        
File "/home/tes/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/ut
il/deprecation.py", line 316, in new_func                                          
return func(*args, **kwargs)                                                   
File "/home/tes/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/framework/importer.py", line 541, in import_graph_def
raise ValueError('No op named %s in defined operations.' % node.op)            
ValueError: No op named NonMaxSuppressionV3 in defined operations.
Could any one help me on this issue?
 
Thanks
  • Up0
  • Down0

Opinions expressed in the content posted here are the personal opinions of the original authors, and do not necessarily reflect those of Qualcomm Incorporated or its subsidiaries (“Qualcomm”). The content is provided for informational purposes only and is not meant to be an endorsement or representation by Qualcomm or any other party. This site may also provide links or references to non-Qualcomm sites and resources. Qualcomm makes no representations, warranties, or other commitments whatsoever about any non-Qualcomm sites or third-party resources that may be referenced, accessible from, or linked to this site.