4.1.3.3. Other Algorithm Model Samples

Other algorithm model samples refer to the samples in the 05_miscellaneous/ and 06_custom_op/ folders. Amongst, the 05_miscellaneous/ folder contains miscellaneous samples used for other features in the X3 Toolchain, e.g. how to enable the RGB data trained model receives YUV data at Runtime using D-Robotics’ model conversion Toolchain. The 06_custom_op/ folder is a user custom OP sample helps developers understand how to add custom OP when there are Toolchain unsupported OPs.

Usages and directories of the above mentioned samples please refer to below contents.

4.1.3.3.1. How to Use

Each sample is encapsulated into a shell script in order. Run shell scripts in order to understand usage of corresponding tools, API features and usage etc.

4.1.3.3.2. Miscellaneous Samples

4.1.3.3.2.1. lenet_gray

4.1.3.3.2.1.1. Directory

01_lenet_gray/
.
└── mapper
    ├── 00_init.sh                  // Obtain the model and calibration dataset required for the current sample
    ├── 01_check.sh                 // Check model validity
    ├── 02_get_mnist.sh             // Obtain the calibration dataset
    ├── 03_build.sh                 // Convert the debugging model
    ├── 04_inference.sh             // Run inference using the onnx runtime
    ├── README.md
    ├── lenet_gray_config.yaml
    ├── postprocess.py
    ├── preprocess.py
    └── process_mnist.py

4.1.3.3.2.1.2. Description

This section contains model conversion, simulator runtime and on-board runtime samples of gray image models.

4.1.3.3.2.2. resnet50_feature

4.1.3.3.2.2.1. Directory

02_resnet50_feature
.
└── mapper
    ├── 00_init.sh                  // Obtain the model and calibration dataset required for the current sample
    ├── 01_check.sh                 // Check model validity
    ├── 02_build.sh                 // Convert the debugging model
    ├── 03_inference.sh             // Run inference using the onnx runtime
    ├── README.md
    ├── inference.py
    └── resnet50_feature_config.yaml

4.1.3.3.2.2.2. Description

This section contains model conversion, simulator runtime and on-board runtime samples of resnet50_feature.

4.1.3.3.2.3. vector-diff

4.1.3.3.2.3.1. Directory

03_vector_diff
.
└── mobilenet_mapper
    ├── 01_inference_rt.sh
    ├── 02_vec_diff.sh                  // Analyze output data using the vec_diff tool
    ├── ILSVRC2012_val_00000001.bin
    └── README.md

4.1.3.3.2.3.2. Output List

Some CSV list(s) whose filename(s) is/are specified by the vec_diff -o command. In which: Left Files, Right Files, Cosine Similarity,Relative Euclidean Distance, Max Absolute Error and Mean Square Error. See below:

Left Files

Right Files

Cosine Similarity

Relative Euclidean Distance

Max Absolute Error

Mean Square Error

Layerxxx-quanti-input.txt

Layerxxx-float-input.txt

xxx

xxx

xxx

xxx

Layerxxx-quanti-param.txt

Layerxxx-float-param.txt

xxx

xxx

xxx

xxx

4.1.3.3.2.4. multi_input_example

4.1.3.3.2.4.1. Directory

04_multi_input_example
.
└── mapper
    ├── 00_init.sh                  // Obtain the model and calibration dataset required for the current sample
    ├── 01_check.sh                 // Check model validity
    ├── 02_preprocess.sh            // Run model preprocess
    ├── 03_build.sh                 // Convert the debugging model
    ├── 04_inference.sh             // Inference a single image
    ├── README.md
    ├── data_preprocess.py
    ├── data_transformer.py
    ├── inference.py
    └── mobilenetv2_multi_config.yaml

4.1.3.3.2.4.2. Description

This section contains model conversion, simulator runtime and on-board runtime samples of multi-input models.

4.1.3.3.2.5. model_verifier

4.1.3.3.2.5.1. Directory

07_model_verifier
.
├── 00_init.sh                      // Obtain the model and calibration dataset required for the current sample
├── 01_preprocess.sh                // Run model preprocess
├── 02_build.sh                     // Convert the debugging model
├── 03_model_verify.sh              // Run model verification
├── calibration_data_feature
├── preprocess.py
├── README.md
├── mobilenet_config_bgr.yaml
├── mobilenet_config_yuv444.yaml
└── resnet50_featuremap_config.yaml

4.1.3.3.2.5.2. Description

This section contains samples of model verification tool.

4.1.3.3.2.6. model_info

4.1.3.3.2.6.1. Directory

08_model_info
.
├── 00_init.sh                      // Obtain the model and calibration dataset required for the current sample
├── 01_preprocess.sh                // Run model preprocess
├── 02_build.sh                     // Convert the debugging model
├── 03_model_info_check.sh          // Obtain and print model information
├── README.md
├── mobilenet_config.yaml
└── preprocess.py

4.1.3.3.2.6.2. Description

This section contains a samples of the model validation tool.

4.1.3.3.2.7. mobilenet_bgr

4.1.3.3.2.7.1. Directory

09_mobilenet_bgr
.
└── mapper
    ├── 00_init.sh                // Obtain the model and calibration dataset required for the current sample
    ├── 01_check.sh               // Check model validity
    ├── 02_preprocess.sh          // Preprocess the dataset
    ├── 03_build.sh               // Convert the debugging model
    ├── 04_inference.sh           // Inference a single image
    ├── README.md
    ├── mobilenet_config.yaml
    ├── postprocess.py
    └── preprocess.py

4.1.3.3.2.7.2. Description

This is a sample of the MobileNetv1 model whose input_type_rt is specified as bgr.

4.1.3.3.2.8. mobilenet_yuv444

4.1.3.3.2.8.1. Directory

10_mobilenet_yuv444
.
└── mapper
    ├── 00_init.sh                // Obtain the model and calibration dataset required for the current sample
    ├── 01_check.sh               // Check model validity
    ├── 02_preprocess.sh          // Preprocess the dataset
    ├── 03_build.sh               // Convert the debugging model
    ├── 04_inference.sh           // Inference a single image
    ├── README.md
    ├── mobilenet_config.yaml
    ├── postprocess.py
    └── preprocess.py

4.1.3.3.2.8.2. Description

This is a sample of the MobileNetv1 model whose input_type_rt is specified as yuv444.

4.1.3.3.3. User Custom OP Sample

4.1.3.3.3.1. Directory

06_custom_op
.
└── mapper
    ├── 00_init.sh
    ├── 03_build.sh
    ├── 04_inference.sh
    ├── README.md
    ├── create_onnx.py
    ├── custom_op_config.yaml
    ├── custom_op_inference.py
    └── horizon_ops.py

4.1.3.3.3.2. Description

When converting the open-source framework trained floating-point model into fixed-point model, the conversion will fail if there are Toolchain unsupported OP(s). In such case, developers can still convert the model by adding custom OP using the custom OP feature.

The mapper folder contains the scripts and configuration file required by running this sample.