This document describes the process of setting up and running the Arm® Ethos™-U NPU Image Classification example.
This use-case example solves the classical computer vision problem of image classification. The ML sample was developed using the MobileNet v2 model that was trained on the ImageNet dataset.
Use-case code could be found in the following directory:source/use_case/img_class.
In addition to the already specified build option in the main documentation, the Image Classification use-case specifies:
img_class_MODEL_TFLITE_PATH - The path to the NN model file in the
TFLite format. The model is then processed and included in the application
axf file. The default value points to one of the delivered set of models.
Note that the parameters
ETHOS_U_NPU_ENABLED must be aligned with the chosen model. In other words:
ETHOS_U_NPU_ENABLEDis set to
1, then the NN model is assumed to be optimized. The model naturally falls back to the Arm® Cortex®-M CPU if an unoptimized model is supplied.
ETHOS_U_NPU_ENABLEDis set to
0, the NN model is assumed to be unoptimized. Supplying an optimized model in this case results in a runtime error.
img_class_FILE_PATH: The path to the directory containing the images, or a path to a single image file, that is to be used in the application. The default value points to the
resources/img_class/samples folder containing the delivered set of images.
For further information, please refer to: Add custom input data section.
img_class_IMAGE_SIZE: The NN model requires input images to be of a specific size. This parameter defines the size of the image side in pixels. Images are considered squared. The default value is
224, which is what the supplied MobilenetV2-1.0 model expects.
img_class_LABELS_TXT_FILE: The path to the text file for the label. The file is used to map a classified class index to the text label. The default value points to the delivered
labels.txt file inside the delivery package. Change this parameter to point to the custom labels file to map custom NN model output correctly.
img_class_ACTIVATION_BUF_SZ: The intermediate, or activation, buffer size reserved for the NN model. By default, it is set to 2MiB and is enough for most models.
USE_CASE_BUILD: is set to
img_class to only build this example.
To build ONLY the Image Classification example application, add
-DUSE_CASE_BUILD=img_class to the
cmake command line, as specified in: Building.
Note: This section describes the process for configuring the build for the MPS3: SSE-300. To build for a different target platform, please refer to: Building.
Create a build directory and navigate inside, like so:
mkdir build_img_class && cd build_img_class
On Linux, when providing only the mandatory arguments for the CMake configuration, execute the following command to build only Image Classification application to run on the Ethos-U55 Fast Model:
cmake ../ -DUSE_CASE_BUILD=img_class
To configure a build that can be debugged using Arm DS specify the build type as
Debug and then use the
Arm Compiler toolchain file:
cmake .. \ -DCMAKE_TOOLCHAIN_FILE=scripts/cmake/toolchains/bare-metal-armclang.cmake \ -DCMAKE_BUILD_TYPE=Debug \ -DUSE_CASE_BUILD=img_class
For further information, please refer to:
Note: If re-building with changed parameters values, we recommend that you clean the build directory and re-run the CMake command.
If the CMake command succeeds, build the application as follows:
To see compilation and link details, add
Results of the build are placed under the
build/bin folder, like so:
bin ├── ethos-u-img_class.axf ├── ethos-u-img_class.htm ├── ethos-u-img_class.map └── sectors ├── images.txt └── img_class ├── ddr.bin └── itcm.bin
bin folder contains the following files:
ethos-u-img_class.axf: The built application binary for the Image Classification use-case.
ethos-u-img_class.map: Information from building the application. For example: The libraries used, what was optimized, and the location of objects.
ethos-u-img_class.htm: Human readable file containing the call graph of application functions.
sectors/img_class: Folder containing the built application. It is split into files for loading into different FPGA memory regions.
sectors/images.txt: Tells the FPGA which memory regions to use for loading the binaries in the
The application anomaly detection is set up to perform inferences on data found in the folder, or an individual file, that is pointed to by the parameter
To run the application with your own images, first create a folder to hold them and then copy the custom images into the following folder:
mkdir /tmp/custom_images cp custom_image1.bmp /tmp/custom_images/
Note: Clean the build directory before re-running the CMake command.
img_class_FILE_PATH to the location of this folder when building:
cmake .. \ -Dimg_class_FILE_PATH=/tmp/custom_images/ \ -DUSE_CASE_BUILD=img_class
The images found in the
img_class_FILE_PATH folder are picked up and automatically converted to C++ files during the CMake configuration stage. They are then compiled into the application during the build phase for performing inference with.
The log from the configuration stage tells you what image directory path has been used:
-- User option img_class_FILE_PATH is set to /tmp/custom_images -- User option img_class_IMAGE_SIZE is set to 224 ... -- Generating image files from /tmp/custom_images ++ Converting custom_image1.bmp to custom_image1.cc ... -- Defined build user options: ... -- img_class_FILE_PATH=/tmp/custom_images -- img_class_IMAGE_SIZE=224
After compiling, your custom images have now replaced the default ones in the application.
Note: The CMake parameter
IMAGE_SIZEmust match the model input size. When building the application, if the size of any image does not match
IMAGE_SIZE, then it is rescaled and padded so that it does.
The application performs inference using the model pointed to by the CMake parameter
Note: If you want to run the model using an Ethos-U, ensure that your custom model has been successfully run through the Vela compiler before continuing.
For further information: Optimize model with Vela compiler.
To run the application with a custom model, you must provide a
labels_<model_name>.txt file of labels that are associated with the model. Each line of the file must correspond to one of the outputs in your model.
Refer to the provided
labels_mobilenet_v2_1.0_224.txt file for an example.
Then, you must set
img_class_MODEL_TFLITE_PATH to the location of the Vela processed model file and
img_class_LABELS_TXT_FILE to the location of the associated labels file.
cmake .. \ -Dimg_class_MODEL_TFLITE_PATH=<path/to/custom_model_after_vela.tflite> \ -Dimg_class_LABELS_TXT_FILE=<path/to/labels_custom_model.txt> \ -DUSE_CASE_BUILD=img_class
Note: Clean the build directory before re-running the CMake command.
.tflite model file pointed to by
img_class_MODEL_TFLITE_PATH, and the labels text file pointed to by
img_class_LABELS_TXT_FILE are converted to C++ files during the CMake configuration stage. They are then compiled into the application for performing inference with.
The log from the configuration stage tells you what model path and labels file have been used, for example:
-- User option img_class_MODEL_TFLITE_PATH is set to <path/to/custom_model_after_vela.tflite> ... -- User option img_class_LABELS_TXT_FILE is set to <path/to/labels_custom_model.txt> ... -- Using <path/to/custom_model_after_vela.tflite> ++ Converting custom_model_after_vela.tflite to\ custom_model_after_vela.tflite.cc -- Generating labels file from <path/to/labels_custom_model.txt> -- writing to <path/to/build/generated/src/Labels.cc> ...
After compiling, your custom model has now replaced the default one in the application.
The FVP is available publicly from Arm Ecosystem FVP downloads.
For the Ethos-U evaluation, please download the MPS3 based version of the Arm® Corstone™-300 model that contains Cortex-M55 and offers a choice of the Ethos-U55 and Ethos-U65 processors.
To install the FVP:
Unpack the archive.
Run the install script in the extracted package:
The pre-built application binary
ethos-u-img_class.axf can be found in the
bin/mps3-sse-300 folder of the delivery package.
Assuming that the install location of the FVP was set to
~/FVP_install_location, then the simulation can be started by using:
A log output appears on the terminal:
telnetterminal0: Listening for serial connection on port 5000 telnetterminal1: Listening for serial connection on port 5001 telnetterminal2: Listening for serial connection on port 5002 telnetterminal5: Listening for serial connection on port 5003
This also launches a telnet window with the standard output of the sample application. It also includes error log entries containing information about the pre-built application version, TensorFlow Lite Micro library version used, and data types. The log also includes the input and output tensor sizes of the model compiled into the executable binary.
After the application has started, if
img_class_FILE_PATH points to a single file, or even a folder that contains a single image, then the inference starts immediately. If there are multiple inputs, it outputs a menu and then waits for input from the user:
User input required Enter option number from: 1. Classify next ifm 2. Classify ifm at chosen index 3. Run classification on all ifm 4. Show NN model info 5. List ifm Choice:
What the preceding choices do:
Classify next image: Runs a single inference on the next in line image from the collection of the compiled images.
Classify image at chosen index: Runs inference on the chosen image.
Note: Please make sure to select image index from within the range of supplied audio clips during application build. By default, a pre-built application has four images, with indexes from
Run classification on all images: Triggers sequential inference executions on all built-in images.
Show NN model info: Prints information about the model data type, input, and output, tensor sizes:
INFO - Model info: INFO - Model INPUT tensors: INFO - tensor type is INT8 INFO - tensor occupies 150528 bytes with dimensions INFO - 0: 1 INFO - 1: 224 INFO - 2: 224 INFO - 3: 3 INFO - Quant dimension: 0 INFO - Scale = 0.007843 INFO - ZeroPoint = -1 INFO - Model OUTPUT tensors: INFO - tensor type is INT8 INFO - tensor occupies 1001 bytes with dimensions INFO - 0: 1 INFO - 1: 1001 INFO - Quant dimension: 0 INFO - Scale = 0.03906 INFO - ZeroPoint = -128 INFO - Activation buffer (a.k.a tensor arena) size used: 1510004 INFO - Number of operators: 1 INFO - Operator 0: ethos-u
List Images: Prints a list of pair image indexes. The original filenames are embedded in the application, like so:
INFO - List of Files: INFO - 0 => cat.bmp INFO - 1 => dog.bmp INFO - 2 => kimono.bmp INFO - 3 => tiger.bmp
Please select the first menu option to execute Image Classification.
The following example illustrates an application output for classification:
INFO - Running inference on image 0 => cat.bmp INFO - Final results: INFO - Total number of inferences: 1 INFO - 0) 282 (0.753906) -> tabby, tabby cat INFO - 1) 286 (0.148438) -> Egyptian cat INFO - 2) 283 (0.062500) -> tiger cat INFO - 3) 458 (0.003906) -> bow tie, bow-tie, bowtie INFO - 4) 288 (0.003906) -> lynx, catamount INFO - Profile for Inference: INFO - NPU AXI0_RD_DATA_BEAT_RECEIVED beats: 2468259 INFO - NPU AXI0_WR_DATA_BEAT_WRITTEN beats: 1151319 INFO - NPU AXI1_RD_DATA_BEAT_RECEIVED beats: 432351 INFO - NPU ACTIVE cycles: 7345741 INFO - NPU IDLE cycles: 431 INFO - NPU TOTAL cycles: 7346172
It can take several minutes to complete one inference run. The average time is around 2-3 minutes.
The log shows the inference results for
image 0, so
index, that corresponds to
cat.bmp in the sample image resource folder.
The profiling section of the log shows that for this inference:
Ethos-U PMU report:
7,346,172 total cycle: The number of NPU cycles.
7,345,741 active cycles: The number of NPU cycles that were used for computation.
413 idle cycles: The number of cycles for which the NPU was idle.
2,468,259 AXI0 read beats: The number of AXI beats with read transactions from AXI0 bus. AXI0 is the bus where the Ethos-U NPU reads and writes to the computation buffers, activation buf, or tensor arenas.
1,151,319 AXI0 write beats: The number of AXI beats with write transactions to AXI0 bus.
432,351 AXI1 read beats: The number of AXI beats with read transactions from AXI1 bus. AXI1 is the bus where the Ethos-U NPU reads the model. So, read-only.
For FPGA platforms, a CPU cycle count can also be enabled. However, do not use cycle counters for FVP, as the CPU model is not cycle-approximate or cycle-accurate
The application prints the top five classes with indexes, a confidence score, and labels from the associated labels_mobilenet_v2_1.0_224.txt file. The FVP window also shows the output on its LCD section.