我已经通过 ssd inception v2 模型训练了一个模型，并使用“export_inference_graph.py”生成了frozen_inference_graph.pb
我尝试使用以下命令生成模型，然后尝试使用 mvNCCompiler 将其转换为 movidius 图
python3 /home/chanchihang/Special_Topic/models/research/object_detection/export_inference_graph.py
--input_type image_tensor
--pipeline_config_path /home/chanchihang/Special_Topic/ssd_inception_v2_coco.config
--trained_checkpoint_prefix /home/chanchihang/Special_Topic/models/research/inceptionV2_Training/model.ckpt-20000
--output_directory /home/chanchihang/Special_Topic/models/research/inceptionV2_Training/
然后我尝试通过 mvNCCompiler 将 model.ckpt.meta 转换为 movidius 图并得到下面的错误
mvNCCompile model.ckpt.meta -in=image_tensor -on=detection_scores -is 224 224 -o test.graph
Traceback (most recent call last):
File "/usr/local/bin/mvNCCompile", line 118, in <module>
create_graph(args.network, args.inputnode, args.outputnode, args.outfile, args.nshaves, args.inputsize, args.weights)
File "/usr/local/bin/mvNCCompile", line 104, in create_graph
net = parse_tensor(args, myriad_config)
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 961, in numpy.random.mtrand.RandomState.uniform
File "common.pyx", line 557, in numpy.random.common.cont
TypeError: 'NoneType' object cannot be interpreted as an integer