[IE Samples Docs] Add rows with model download and conversion (#6626)
* Add information about model preparation * Add the missed file extension for the onnx model * [c] Add information about model preparation * [cpp] Add information about model preparation * Update the string about model converter * Specify a model name at steps 1 and 2 of the running example
This commit is contained in:
Dmitry Pigasin authored and 
GitHub committed
1 parent
2378593bd5
commit
1cfe27ae1f
20 files changed
+399
-239
No files matched your search
@@ -60,8 +60,8 @@ To run the sample, you need specify a model and image:
|
||||
|
||||
Running the application with the `-h` option yields the following usage message:
|
||||
|
||||
```sh
|
||||
./classification_sample_async -h
|
||||
```
|
||||
<path_to_sample>/classification_sample_async -h
|
||||
InferenceEngine:
|
||||
API version ............ <version>
|
||||
Build .................. <build>
|
||||
@@ -85,33 +85,43 @@ Options:
|
||||
|
||||
Running the application with the empty list of options yields the usage message given above and an error message.
|
||||
|
||||
You can do inference of an image using a trained [AlexNet network](https://docs.openvinotoolkit.org/latest/omz_models_model_alexnet.html) on GPU using the following command:
|
||||
### Example
|
||||
1. Download a pre-trained model using [Model Downloader](@ref omz_tools_downloader_README):
|
||||
```
|
||||
python <path_to_omz_tools>/downloader.py --name alexnet
|
||||
```
|
||||
|
||||
```sh
|
||||
./classification_sample_async -m <path_to_model>/alexnet_fp32.xml -i <path_to_image>/cat.bmp -d GPU
|
||||
2. If a model is not in the Inference Engine IR or ONNX format, it must be converted. You can do this using the model converter script:
|
||||
|
||||
```
|
||||
python <path_to_omz_tools>/converter.py --name alexnet
|
||||
```
|
||||
|
||||
3. Perform inference of `car.bmp` using `alexnet` model on a `GPU`, for example:
|
||||
|
||||
```
|
||||
<path_to_sample>/classification_sample_async -m <path_to_model>/alexnet.xml -i <path_to_image>/car.bmp -d GPU
|
||||
```
|
||||
|
||||
## Sample Output
|
||||
|
||||
By default the application outputs top-10 inference results for each infer request.
|
||||
|
||||
```sh
|
||||
classification_sample_async -m alexnet_fp32/alexnet.xml -i car_1.bmp -d GPU
|
||||
```
|
||||
[ INFO ] InferenceEngine:
|
||||
API version ............
|
||||
Build ..................
|
||||
Description ....... API
|
||||
[ INFO ] Parsing input parameters
|
||||
IE version ......... 2021.4.0
|
||||
Build ........... 2021.4.0-3839-cd81789d294-releases/2021/4
|
||||
[ INFO ] Parsing input parameters
|
||||
[ INFO ] Files were added: 1
|
||||
[ INFO ] car_1.bmp
|
||||
[ INFO ] C:\images\car.bmp
|
||||
[ INFO ] Loading Inference Engine
|
||||
[ INFO ] Device info:
|
||||
GPU
|
||||
clDNNPlugin version ......... <version>
|
||||
Build ........... <build>
|
||||
clDNNPlugin version ......... 2021.4.0
|
||||
Build ........... 2021.4.0-3839-cd81789d294-releases/2021/4
|
||||
|
||||
[ INFO ] Loading network files:
|
||||
alexnet_fp32/alexnet.xml
|
||||
[ INFO ] C:\openvino\deployment_tools\open_model_zoo\tools\downloader\public\alexnet\FP32\alexnet.xml
|
||||
[ INFO ] Preparing input blobs
|
||||
[ WARNING ] Image is resized from (749, 637) to (227, 227)
|
||||
[ INFO ] Batch size is 1
|
||||
@@ -132,20 +142,20 @@ classification_sample_async -m alexnet_fp32/alexnet.xml -i car_1.bmp -d GPU
|
||||
|
||||
Top 10 results:
|
||||
|
||||
Image car_1.bmp
|
||||
Image C:\images\car.bmp
|
||||
|
||||
classid probability
|
||||
------- -----------
|
||||
656 0.5491584
|
||||
874 0.1101241
|
||||
654 0.0559816
|
||||
436 0.0488046
|
||||
581 0.0330480
|
||||
705 0.0307707
|
||||
734 0.0185521
|
||||
627 0.0162536
|
||||
675 0.0145008
|
||||
757 0.0125437
|
||||
656 0.6645315
|
||||
654 0.1121185
|
||||
581 0.0698451
|
||||
874 0.0334973
|
||||
436 0.0259718
|
||||
817 0.0173190
|
||||
675 0.0109321
|
||||
511 0.0109075
|
||||
569 0.0083093
|
||||
717 0.0063173
|
||||
|
||||
[ INFO ] Execution successful
|
||||
|
||||
|
||||
@@ -47,33 +47,45 @@ To run the sample, you need specify a model and image:
|
||||
>
|
||||
> - The sample accepts models in ONNX format (.onnx) that do not require preprocessing.
|
||||
|
||||
You can do inference of an image using a trained AlexNet network on a GPU using the following command:
|
||||
### Example
|
||||
1. Download a pre-trained model using [Model Downloader](@ref omz_tools_downloader_README):
|
||||
```
|
||||
python <path_to_omz_tools>/downloader.py --name alexnet
|
||||
```
|
||||
|
||||
```sh
|
||||
./hello_classification <path_to_model>/alexnet_fp32.xml <path_to_image>/car.png GPU
|
||||
2. If a model is not in the Inference Engine IR or ONNX format, it must be converted. You can do this using the model converter script:
|
||||
|
||||
```
|
||||
python <path_to_omz_tools>/converter.py --name alexnet
|
||||
```
|
||||
|
||||
3. Perform inference of `car.bmp` using `alexnet` model on a `GPU`, for example:
|
||||
|
||||
```
|
||||
<path_to_sample>/hello_classification <path_to_model>/alexnet.xml <path_to_image>/car.bmp GPU
|
||||
```
|
||||
|
||||
## Sample Output
|
||||
|
||||
The application outputs top-10 inference results.
|
||||
|
||||
```sh
|
||||
```
|
||||
Top 10 results:
|
||||
|
||||
Image /opt/intel/openvino/deployment_tools/demo/car.png
|
||||
Image C:\images\car.bmp
|
||||
|
||||
classid probability
|
||||
------- -----------
|
||||
479 0.7562194
|
||||
511 0.0760387
|
||||
436 0.0724114
|
||||
817 0.0462140
|
||||
656 0.0301230
|
||||
661 0.0056171
|
||||
581 0.0031623
|
||||
468 0.0029917
|
||||
717 0.0023081
|
||||
627 0.0016193
|
||||
656 0.6664789
|
||||
654 0.1129405
|
||||
581 0.0684867
|
||||
874 0.0333845
|
||||
436 0.0261321
|
||||
817 0.0167310
|
||||
675 0.0109796
|
||||
511 0.0105919
|
||||
569 0.0081782
|
||||
717 0.0063356
|
||||
|
||||
This sample is an API example, for any performance measurements please use the dedicated benchmark_app tool
|
||||
```
|
||||
|
||||
@@ -64,17 +64,29 @@ ffmpeg -i cat.jpg -pix_fmt nv12 cat.yuv
|
||||
>
|
||||
> - The sample accepts models in ONNX format (.onnx) that do not require preprocessing.
|
||||
|
||||
You can perform inference on an NV12 image using a trained AlexNet network on CPU with the following command:
|
||||
### Example
|
||||
1. Download a pre-trained model using [Model Downloader](@ref omz_tools_downloader_README):
|
||||
```
|
||||
python <path_to_omz_tools>/downloader.py --name alexnet
|
||||
```
|
||||
|
||||
```sh
|
||||
./hello_nv12_input_classification <path_to_model>/alexnet_fp32.xml <path_to_image>/cat.yuv 300x300 CPU
|
||||
2. If a model is not in the Inference Engine IR or ONNX format, it must be converted. You can do this using the model converter script:
|
||||
|
||||
```
|
||||
python <path_to_omz_tools>/converter.py --name alexnet
|
||||
```
|
||||
|
||||
3. Perform inference of NV12 image using `alexnet` model on a `CPU`, for example:
|
||||
|
||||
```
|
||||
<path_to_sample>/hello_nv12_input_classification <path_to_model>/alexnet.xml <path_to_image>/cat.yuv 300x300 CPU
|
||||
```
|
||||
|
||||
## Sample Output
|
||||
|
||||
The application outputs top-10 inference results.
|
||||
|
||||
```sh
|
||||
```
|
||||
[ INFO ] Files were added: 1
|
||||
[ INFO ] ./cat.yuv
|
||||
Batch size is 1
|
||||
|
||||
@@ -27,8 +27,8 @@ To build the sample, please use instructions available at [Build the Sample Appl
|
||||
|
||||
To see quired information, run the following:
|
||||
|
||||
```sh
|
||||
./hello_query_device -h
|
||||
```
|
||||
<path_to_sample>/hello_query_device -h
|
||||
Usage : hello_query_device
|
||||
```
|
||||
|
||||
@@ -36,8 +36,7 @@ Usage : hello_query_device
|
||||
|
||||
The application prints all available devices with their supported metrics and default values for configuration parameters:
|
||||
|
||||
```sh
|
||||
./hello_query_device
|
||||
```
|
||||
Available devices:
|
||||
Device: CPU
|
||||
Metrics:
|
||||
|
||||
@@ -51,14 +51,26 @@ To run the sample, you need specify a model and image:
|
||||
|
||||
You can use the following command to do inference on CPU of an image using a trained SSD network:
|
||||
|
||||
```sh
|
||||
hello_reshape_ssd <path_to_model> <path_to_image> <device> <batch>
|
||||
```
|
||||
<path_to_sample>/hello_reshape_ssd <path_to_model> <path_to_image> <device> <batch>
|
||||
```
|
||||
|
||||
with one image and [person-detection-retail-0013](https://docs.openvinotoolkit.org/latest/omz_models_intel_person_detection_retail_0013_description_person_detection_retail_0013.html) model
|
||||
### Example
|
||||
1. Download a pre-trained model using [Model Downloader](@ref omz_tools_downloader_README):
|
||||
```
|
||||
python <path_to_omz_tools>/downloader.py --name person-detection-retail-0013
|
||||
```
|
||||
|
||||
```sh
|
||||
hello_reshape_ssd <path_to_model>/person-detection-retail-0013.xml <path_to_image>/inputImage.bmp CPU 1
|
||||
2. `person-detection-retail-0013` model does not need to be converted, because it is already in necessary format, so you can skip this step. If you want to use a other model that is not in the Inference Engine IR or ONNX format, you can convert it using the model converter script:
|
||||
|
||||
```
|
||||
python <path_to_omz_tools>/converter.py --name <model_name>
|
||||
```
|
||||
|
||||
3. Perform inference of `person_detection.png` using `person-detection-retail-0013` model on a `GPU`, for example:
|
||||
|
||||
```
|
||||
<path_to_sample>/hello_reshape_ssd <path_to_model>/person-detection-retail-0013.xml <path_to_image>/person_detection.png GPU 1
|
||||
```
|
||||
|
||||
## Sample Output
|
||||
@@ -67,13 +79,11 @@ The application renders an image with detected objects enclosed in rectangles. I
|
||||
of the detected objects along with the respective confidence values and the coordinates of the
|
||||
rectangles to the standard output stream.
|
||||
|
||||
```sh
|
||||
hello_reshape_ssd person-detection-retail-0013/FP16/person-detection-retail-0013.xml person_detection.png CPU 1
|
||||
|
||||
```
|
||||
Resizing network to the image size = [960x1699] with batch = 1
|
||||
Resulting input shape = [1,3,960,1699]
|
||||
Resulting output shape = [1,1,200,7]
|
||||
[0,1] element, prob = 0.721457, bbox = (852.37,187.54)-(983.326,520.672), batch id = 0
|
||||
[0,1] element, prob = 0.722292, bbox = (852.382,187.756)-(983.352,520.733), batch id = 0
|
||||
The resulting image was saved in the file: hello_reshape_ssd_output.jpg
|
||||
|
||||
This sample is an API example, for any performance measurements please use the dedicated benchmark_app tool
|
||||
|
||||
@@ -50,7 +50,7 @@ To run the sample, you need specify a model wights and ubyte image:
|
||||
|
||||
Running the application with the `-h` option yields the following usage message:
|
||||
|
||||
```sh
|
||||
```
|
||||
ngraph_function_creation_sample -h
|
||||
[ INFO ] InferenceEngine:
|
||||
API version ............<version>
|
||||
@@ -75,8 +75,8 @@ Running the application with the empty list of options yields the usage message
|
||||
|
||||
You can do inference of an image using a pre-trained model on a GPU using the following command:
|
||||
|
||||
```sh
|
||||
./ngraph_function_creation_sample -m <path_to_weights_file>/lenet.bin -i <path_to_image> -d GPU
|
||||
```
|
||||
<path_to_sample>/ngraph_function_creation_sample -m <path_to_weights_file>/lenet.bin -i <path_to_image> -d GPU
|
||||
```
|
||||
|
||||
## Sample Output
|
||||
|
||||
@@ -41,9 +41,9 @@ To run the sample, you need specify a model and image:
|
||||
- you can use [public](@ref omz_models_public_index) or [Intel's](@ref omz_models_intel_index) pre-trained models from the Open Model Zoo. The models can be downloaded using the [Model Downloader](@ref omz_tools_downloader_README).
|
||||
- you can use images from the media files collection available at https://storage.openvinotoolkit.org/data/test_data.
|
||||
|
||||
Running the application with the <code>-h</code> option yields the following usage message:
|
||||
Running the application with the `-h` option yields the following usage message:
|
||||
|
||||
```sh
|
||||
```
|
||||
./object_detection_sample_ssd -h
|
||||
InferenceEngine:
|
||||
API version ............ <version>
|
||||
@@ -75,18 +75,30 @@ Running the application with the empty list of options yields the usage message
|
||||
>
|
||||
> - The sample accepts models in ONNX format (\*.onnx) that do not require preprocessing.
|
||||
|
||||
For example, to do inference on a CPU with the OpenVINO™ toolkit person detection SSD models, run one of the following commands:
|
||||
### Example
|
||||
1. Download a pre-trained model using [Model Downloader](@ref omz_tools_downloader_README):
|
||||
```
|
||||
python <path_to_omz_tools>/downloader.py --name person-detection-retail-0013
|
||||
```
|
||||
|
||||
2. `person-detection-retail-0013` model does not need to be converted, because it is already in necessary format, so you can skip this step. If you want to use a other model that is not in the Inference Engine IR or ONNX format, you can convert it using the model converter script:
|
||||
|
||||
```
|
||||
python <path_to_omz_tools>/converter.py --name <model_name>
|
||||
```
|
||||
|
||||
3. For example, to do inference on a CPU with the OpenVINO™ toolkit person detection SSD models, run one of the following commands:
|
||||
|
||||
- with one image and [person-detection-retail-0013](https://docs.openvinotoolkit.org/latest/omz_models_intel_person_detection_retail_0013_description_person_detection_retail_0013.html) model
|
||||
|
||||
```sh
|
||||
./object_detection_sample_ssd -m <path_to_model>/person-detection-retail-0013.xml -i <path_to_image>/inputImage.bmp -d CPU
|
||||
```
|
||||
<path_to_sample>/object_detection_sample_ssd -m <path_to_model>/person-detection-retail-0013.xml -i <path_to_image>/person_detection.png -d CPU
|
||||
```
|
||||
|
||||
- with one image and [person-detection-retail-0002](https://docs.openvinotoolkit.org/latest/omz_models_intel_person_detection_retail_0002_description_person_detection_retail_0002.html) model
|
||||
|
||||
```sh
|
||||
./object_detection_sample_ssd -m <path_to_model>/person-detection-retail-0002.xml -i <path_to_image>/inputImage.jpg -d GPU
|
||||
```
|
||||
<path_to_sample>/object_detection_sample_ssd -m <path_to_model>/person-detection-retail-0002.xml -i <path_to_image>/person_detection.png -d GPU
|
||||
```
|
||||
|
||||
## Sample Output
|
||||
@@ -95,7 +107,7 @@ The application outputs an image (`out_0.bmp`) with detected objects enclosed in
|
||||
of the detected objects along with the respective confidence values and the coordinates of the
|
||||
rectangles to the standard output stream.
|
||||
|
||||
```sh
|
||||
```
|
||||
object_detection_sample_ssd -m person-detection-retail-0013\FP16\person-detection-retail-0013.xml -i person_detection.png
|
||||
[ INFO ] InferenceEngine:
|
||||
API version ............ <version>
|
||||
|
||||
@@ -87,8 +87,7 @@ To build the sample, please use instructions available at [Build the Sample Appl
|
||||
|
||||
Running the application with the `-h` option yields the following usage message:
|
||||
|
||||
```sh
|
||||
./speech_sample -h
|
||||
```
|
||||
[ INFO ] InferenceEngine:
|
||||
API version ............ <version>
|
||||
Build .................. <number>
|
||||
@@ -132,8 +131,8 @@ Running the application with the empty list of options yields the usage message
|
||||
|
||||
You can use the following model optimizer command to convert a Kaldi nnet1 or nnet2 neural network to Inference Engine Intermediate Representation format:
|
||||
|
||||
```sh
|
||||
python mo.py --framework kaldi --input_model wsj_dnn5b.nnet --counts wsj_dnn5b.counts --remove_output_softmax --output_dir <OUTPUT_MODEL_DIR>
|
||||
```
|
||||
python <path_to_mo>/mo.py --framework kaldi --input_model wsj_dnn5b.nnet --counts wsj_dnn5b.counts --remove_output_softmax --output_dir <OUTPUT_MODEL_DIR>
|
||||
```
|
||||
|
||||
Assuming that the model optimizer (`mo.py`), Kaldi-trained neural network, `wsj_dnn5b.nnet`, and Kaldi class counts file, `wsj_dnn5b.counts`, are in the working directory this produces
|
||||
@@ -151,8 +150,8 @@ All of them can be downloaded from [https://storage.openvinotoolkit.org/models_c
|
||||
|
||||
Once the IR is created, you can use the following command to do inference on Intel® Processors with the GNA co-processor (or emulation library):
|
||||
|
||||
```sh
|
||||
./speech_sample -d GNA_AUTO -bs 2 -i dev93_10.ark -m wsj_dnn5b.xml -o scores.ark -r dev93_scores_10.ark
|
||||
```
|
||||
<path_to_sample>/speech_sample -d GNA_AUTO -bs 2 -i <path_to_ark>/dev93_10.ark -m <path_to_model>/wsj_dnn5b.xml -o scores.ark -r <path_to_ark>/dev93_scores_10.ark
|
||||
```
|
||||
|
||||
Here, the floating point Kaldi-generated reference neural network scores (`dev93_scores_10.ark`) corresponding to the input feature file (`dev93_10.ark`) are assumed to be available
|
||||
@@ -170,7 +169,7 @@ All of them can be downloaded from [https://storage.openvinotoolkit.org/models_c
|
||||
The acoustic log likelihood sequences for all utterances are stored in the file. Example `scores.ark` or `scores.npz`. If the `-r` option is used, a report on the statistical score error is generated for each utterance such as
|
||||
the following:
|
||||
|
||||
```sh
|
||||
```
|
||||
./speech_sample -d GNA_AUTO -bs 2 -i dev93_10.ark -m wsj_dnn5b.xml -o scores.ark -r dev93_scores_10.ark
|
||||
[ INFO ] InferenceEngine:
|
||||
API version ............ <version>
|
||||
|
||||
@@ -40,10 +40,9 @@ To run the sample, you need specify a model and image:
|
||||
- you can use [public](@ref omz_models_public_index) or [Intel's](@ref omz_models_intel_index) pre-trained models from the Open Model Zoo. The models can be downloaded using the [Model Downloader](@ref omz_tools_downloader_README).
|
||||
- you can use images from the media files collection available at https://storage.openvinotoolkit.org/data/test_data.
|
||||
|
||||
Running the application with the <code>-h</code> option yields the following usage message:
|
||||
Running the application with the `-h` option yields the following usage message:
|
||||
|
||||
```sh
|
||||
./style_transfer_sample -h
|
||||
```
|
||||
[ INFO ] InferenceEngine:
|
||||
API version ............<version>
|
||||
Build ..................<build>
|
||||
@@ -65,7 +64,6 @@ Options:
|
||||
-mean_val_b Mean values. Required if the model needs mean values for preprocessing and postprocessing.
|
||||
|
||||
Available target devices: <devices>
|
||||
|
||||
```
|
||||
|
||||
Running the application with the empty list of options yields the usage message given above and an error message.
|
||||
@@ -78,37 +76,47 @@ Running the application with the empty list of options yields the usage message
|
||||
>
|
||||
> - The sample accepts models in ONNX format (\*.onnx) that do not require preprocessing.
|
||||
|
||||
To perform inference of an image using a trained model of fast-neural-style-mosaic-onnx network on Intel® CPUs, use the following command:
|
||||
### Example
|
||||
1. Download a pre-trained model using [Model Downloader](@ref omz_tools_downloader_README):
|
||||
```
|
||||
python <path_to_omz_tools>/downloader.py --name fast-neural-style-mosaic-onnx
|
||||
```
|
||||
|
||||
```sh
|
||||
./style_transfer_sample -i <path_to_image>/cat.bmp -m <path_to_model>/fast-neural-style-mosaic-onnx.onnx
|
||||
2. `fast-neural-style-mosaic-onnx` model does not need to be converted, because it is already in necessary format, so you can skip this step. If you want to use a other model that is not in the Inference Engine IR or ONNX format, you can convert it using the model converter script:
|
||||
|
||||
```
|
||||
python <path_to_omz_tools>/converter.py --name <model_name>
|
||||
```
|
||||
|
||||
3. Perform inference of `car.bmp` and `cat.jpg` using `fast-neural-style-mosaic-onnx` model on a `GPU`, for example:
|
||||
|
||||
```
|
||||
<path_to_sample>/style_transfer_sample -m <path_to_model>/fast-neural-style-mosaic-onnx.onnx -i <path_to_image>/car.bmp <path_to_image>/cat.jpg -d GPU
|
||||
```
|
||||
|
||||
## Sample Output
|
||||
|
||||
The sample application logs each step in a standard output stream and creates an image (`out1.bmp`) or a sequence of images (`out1.bmp`, ..., `out<N>.bmp`) which are redrawn in style of the style transfer model used for the sample.
|
||||
|
||||
```sh
|
||||
style_transfer_sample -m fast-neural-style-mosaic-onnx.onnx -i car.png car_1.bmp
|
||||
```
|
||||
[ INFO ] InferenceEngine:
|
||||
API version ............ <version>
|
||||
Build .................. <build>
|
||||
Description ....... API
|
||||
IE version ......... 2021.4.0
|
||||
Build ........... 2021.4.0-3839-cd81789d294-releases/2021/4
|
||||
[ INFO ] Parsing input parameters
|
||||
[ INFO ] Files were added: 2
|
||||
[ INFO ] car.png
|
||||
[ INFO ] car_1.bmp
|
||||
[ INFO
|
||||
[ INFO ] C:\images\car.bmp
|
||||
[ INFO ] C:\images\cat.jpg
|
||||
[ INFO ] Loading Inference Engine
|
||||
[ INFO ] Device info:
|
||||
CPU
|
||||
MKLDNNPlugin version ......... <version>
|
||||
Build ........... <build>
|
||||
GPU
|
||||
clDNNPlugin version ......... 2021.4.0
|
||||
Build ........... 2021.4.0-3839-cd81789d294-releases/2021/4
|
||||
|
||||
[ INFO ] Loading network files
|
||||
[ INFO ] fast-neural-style-mosaic-onnx.onnx
|
||||
[ INFO ] Loading network files:
|
||||
[ INFO ] C:\openvino\deployment_tools\open_model_zoo\tools\downloader\public\fast-neural-style-mosaic-onnx\fast-neural-style-mosaic-onnx.onnx
|
||||
[ INFO ] Preparing input blobs
|
||||
[ WARNING ] Image is resized from (787, 259) to (224, 224)
|
||||
[ WARNING ] Image is resized from (749, 637) to (224, 224)
|
||||
[ WARNING ] Image is resized from (300, 300) to (224, 224)
|
||||
[ INFO ] Batch size is 2
|
||||
[ INFO ] Preparing output blobs
|
||||
[ INFO ] Loading model to the device
|
||||
|
||||
Reference in new issue
Block a user