OpenVINO之七：转换Caffe模型

1 OpenVINO转换caffe模型步骤

进入目录

C:\Program Files (x86)\IntelSWTools\openvino\deployment_tools\model_optimizer

python mo_caffe.py -h

参数如下：
–input_proto：prototxt文件所在位置
-k：CustomLayersMapping.xml文件所在位置
–input_mode：
–mean_file：
–mean_file_offsets：

–output_dir
–scale SCALE
–log_level {CRITICAL,ERROR,WARN,WARNING,INFO,DEBUG,NOTSET}]
–input
–mean_values
–data_type {FP16,FP32,half,float}] [–disable_fusing]
–disable_resnet_optimization
–finegrain_fusing
–enable_concat_optimization
–extensions
–silent
–freeze_placeholder_with_value
–generate_deprecated_IR_V2

–mean_file_offsets
–disable_omitting_optional
–enable_flattening_nested_params

2 OpenVINO支持的caffe模型

2-1 Classification models:

• AlexNet
• VGG-16, VGG-19
• SqueezeNet v1.0, SqueezeNet v1.1
• ResNet-50, ResNet-101, Res-Net-152
• Inception v1, Inception v2, Inception v3, Inception v4
• CaffeNet
• MobileNet
• Squeeze-and-Excitation Networks: SE-BN-Inception, SE-Resnet-101, SE-ResNet-152, SE-ResNet-50, SE-ResNeXt-101, SE-ResNeXt-50
• ShuffleNet v2

2-2 Object detection models:

• SSD300-VGG16, SSD500-VGG16
• Faster-RCNN
• RefineDet (Myriad plugin only)

2-3 Face detection models:

• VGG Face

2-4 Semantic segmentation models:

• FCN8

3 OpenVINO支持的caffe层与其在Intermediate Representation (IR)中的对应关系

NUMBER	LAYER NAME IN CAFFE*	LAYER NAME IN THE INTERMEDIATE REPRESENTATION
1	Input	Input
2	GlobalInput	Input
3	InnerProduct	FullyConnected
4	Dropout	Ignored, does not appear in IR
5	Convolution	Convolution
6	Deconvolution	Deconvolution
7	Pooling	Pooling
8	BatchNorm	BatchNormalization
9	LRN	Norm
10	Power	Power
11	ReLU	ReLU
12	Scale	ScaleShift
13	Concat	Concat
14	Eltwise	Eltwise
15	Flatten	Flatten
16	Reshape	Reshape
17	Slice	Slice
18	Softmax	SoftMax
19	Permute	Permute
20	ROIPooling	ROIPooling
21	Tile	Tile
22	ShuffleChannel	Reshape + Split + Permute + Concat
23	Axpy	ScaleShift + Eltwise
24	BN	ScaleShift
25	DetectionOutput	DetectionOutput
26	StridedSlice	StridedSlice
27	Bias	Eltwise(operation = sum)

参考资料：
1 Converting a Caffe* Model
2 Supported Framework Layers