C++ Tools

Simple C++ implementations for the client API’s are provided in tools/aiaa folder. These tools are built and installed as part of the source build. Or you can download it directly from the release page.

List Models

Provides implementation for nvidia::aiaa::Client::model() API. For more details refer aiaa-model.cpp

Following are the options available

Option	Description	Default	Example
-h	Prints the help information
-server	Server URI for AIAA Server		-server http://0.0.0.0:5000
-label	Label Name for matching		-label liver
-type	Find Matching Model of type (segmentation/annotation)		-type segmentation
-output	Save output result into a file		-output models.json

Example

nvidiaAIAAListModels \
   -server http://0.0.0.0:5000 \
   -label spleen

Session

Provides implementation for nvidia::aiaa::Client::create_session(), nvidia::aiaa::Client::get_session() and nvidia::aiaa::Client::close_session() APIs. For more details refer aiaa-session.cpp

Following are the options available

Example

nvidiaAIAASession \
   -server http://0.0.0.0:5000 \
   -op create \
   -image _image.nii.gz

nvidiaAIAASession \
   -server http://0.0.0.0:5000 \
   -op delete \
   -session "9ad970be-530e-11ea-84e3-0242ac110007"

DEXTR3D (Annotation)

Provides implementation for nvidia::aiaa::Client::dextra3d() API. For more details refer aiaa-dextra3d.cpp

Following are the options available

Example

nvidiaAIAADEXTR3D \
   -server http://0.0.0.0:5000 \
   -label spleen \
   -points `cat ../test/data/pointset.json` \
   -image _image.nii.gz \
   -output tmp_out.nii.gz \
   -pad 20 \
   -roi 128x128x128 \
   -sigma 3

#(using model instead of label)

nvidiaAIAADEXTR3D \
   -server http://0.0.0.0:5000 \
   -model annotation_ct_spleen \
   -points `cat ../test/data/pointset.json` \
   -image _image.nii.gz \
   -output tmp_out.nii.gz \
   -pad 20 \
   -roi 128x128x128 \
   -sigma 3

Segmentation

Provides implementation for nvidia::aiaa::Client::segmentation() API. For more details refer aiaa-segmentation.cpp

Following are the options available

Example

nvidiaAIAASegmentation \
   -server http://0.0.0.0:5000 \
   -label spleen \
   -image _image.nii.gz \
   -output tmp_out.nii.gz

#(using model instead of label)

nvidiaAIAASegmentation \
   -server http://0.0.0.0:5000 \
   -model segmentation_spleen \
   -image _image.nii.gz \
   -output tmp_out.nii.gz

Deepgrow

Provides implementation for nvidia::aiaa::Client::deepgrow() API. For more details refer aiaa-deepgrow.cpp

Following are the options available

Example

nvidiaAIAADeepgrow \
   -server http://0.0.0.0:5000 \
   -model clara_deepgrow \
   -image _image.nii.gz \
   -fpoints [[283,204,105]] \
   -output tmp_out.nii.gz

nvidiaAIAADeepgrow \
   -server http://0.0.0.0:5000 \
   -model clara_deepgrow \
   -session "9ad970be-530e-11ea-84e3-0242ac110007" \
   -fpoints [[283,204,105]] \
   -output tmp_out.nii.gz

Inference

Provides implementation for nvidia::aiaa::Client::inference() API. For more details refer aiaa-inference.cpp

Following are the options available

Example

nvidiaAIAAInference \
   -server http://0.0.0.0:5000 \
   -model clara_xray_classification_chest_amp \
   -image AC07112011207153620_v2.png \
   -params {}

Mask To Polygon

Provides implementation for nvidia::aiaa::Client::mask2Polygon() API. For more details refer aiaa-mask-polygon.cpp

Following are the options available

Example

nvidiaAIAAMaskPolygon \
   -server http://0.0.0.0:5000 \
   -image tmp_out.nii.gz \
   -output polygonlist.json

Fix Polygon

Provides implementation for nvidia::aiaa::Client::mask2Polygon() API. For more details refer aiaa-fix-polygon.cpp

Following are the options available

Example

nvidiaAIAAFixPolygon \
   -server http://0.0.0.0:5000 \
   -dim 2
   -neighbor 1 \
   -poly `cat ../test/data/polygons.json` \
   -pindex 0 \
   -vindex 17 \
   -xoffset 8 \
   -yoffset 5 \
   -image ../test/data/image_slice_2D.png \
   -output updated_image_2D.png

nvidiaAIAAFixPolygon \
   -server http://0.0.0.0:5000 \
   -dim 3
   -neighbor 1 \
   -neighbor3d 1 \
   -poly `cat ../test/data/polygons3d.json` \
   -sindex 104 \
   -pindex 0 \
   -vindex 4 \
   -xoffset 8 \
   -yoffset 5 \
   -image _image.nii.gz \
   -output image_mask.nii.gz