fix avh demo (#6836)

da0177b4 · Guanghua Yu · GitHub · 847c4c58 · da0177b4 · da0177b4
9 changed file
--- a/deploy/third_engine/demo_avh/.gitignore
+++ b/deploy/third_engine/demo_avh/.gitignore
+include/inputs.h
+include/outputs.h
+__pycache__/
+build/
\ No newline at end of file
--- a/deploy/third_engine/demo_avh/Makefile
+++ b/deploy/third_engine/demo_avh/Makefile
@@ -81,13 +81,27 @@ ${BUILD_DIR}/libcmsis_startup.a: $(CMSIS_STARTUP_SRCS)
 	$(QUIET)$(AR) -cr $(abspath $(BUILD_DIR)/libcmsis_startup.a) $(abspath $(BUILD_DIR))/libcmsis_startup/*.o
 	$(QUIET)$(RANLIB) $(abspath $(BUILD_DIR)/libcmsis_startup.a)
+CMSIS_SHA_FILE=${CMSIS_PATH}/977abe9849781a2e788b02282986480ff4e25ea6.sha
+ifneq ("$(wildcard $(CMSIS_SHA_FILE))","")
+${BUILD_DIR}/cmsis_nn/Source/libcmsis-nn.a:
+	$(QUIET)mkdir -p $(@D)
+	$(QUIET)cd $(CMSIS_PATH)/CMSIS/NN && $(CMAKE) -B $(abspath $(BUILD_DIR)/cmsis_nn) $(CMSIS_NN_CMAKE_FLAGS)
+	$(QUIET)cd $(abspath $(BUILD_DIR)/cmsis_nn) && $(MAKE) all
+else
 # Build CMSIS-NN
 ${BUILD_DIR}/cmsis_nn/Source/SoftmaxFunctions/libCMSISNNSoftmax.a:
 	$(QUIET)mkdir -p $(@D)
 	$(QUIET)cd $(CMSIS_PATH)/CMSIS/NN && $(CMAKE) -B $(abspath $(BUILD_DIR)/cmsis_nn) $(CMSIS_NN_CMAKE_FLAGS)
-	$(QUIET)cd $(abspath $(BUILD_DIR)/cmsis_nn) && $(MAKE) all	
+	$(QUIET)cd $(abspath $(BUILD_DIR)/cmsis_nn) && $(MAKE) all
+endif
 # Build demo application
+ifneq ("$(wildcard $(CMSIS_SHA_FILE))","")
+$(BUILD_DIR)/demo: $(DEMO_MAIN) $(UART_SRCS) $(BUILD_DIR)/stack_allocator.o $(BUILD_DIR)/crt_backend_api.o \
+	${BUILD_DIR}/libcodegen.a ${BUILD_DIR}/libcmsis_startup.a ${BUILD_DIR}/cmsis_nn/Source/libcmsis-nn.a
+	$(QUIET)mkdir -p $(@D)
+	$(QUIET)$(CC) $(PKG_CFLAGS) $(FREERTOS_FLAGS) -o $@ -Wl,--whole-archive $^ -Wl,--no-whole-archive $(PKG_LDFLAGS)
+else
 $(BUILD_DIR)/demo: $(DEMO_MAIN) $(UART_SRCS) $(BUILD_DIR)/stack_allocator.o $(BUILD_DIR)/crt_backend_api.o \
       ${BUILD_DIR}/libcodegen.a ${BUILD_DIR}/libcmsis_startup.a \
       ${BUILD_DIR}/cmsis_nn/Source/SoftmaxFunctions/libCMSISNNSoftmax.a \
@@ -102,6 +116,7 @@ $(BUILD_DIR)/demo: $(DEMO_MAIN) $(UART_SRCS) $(BUILD_DIR)/stack_allocator.o $(BU
       ${BUILD_DIR}/cmsis_nn/Source/PoolingFunctions/libCMSISNNPooling.a
 	$(QUIET)mkdir -p $(@D)
 	$(QUIET)$(CC) $(PKG_CFLAGS) $(FREERTOS_FLAGS) -o $@ -Wl,--whole-archive $^ -Wl,--no-whole-archive $(PKG_LDFLAGS)
+endif
 clean:
 	$(QUIET)rm -rf $(BUILD_DIR)/codegen

--- a/deploy/third_engine/demo_avh/README.md
+++ b/deploy/third_engine/demo_avh/README.md
@@ -5,25 +5,30 @@
 <!--- to you under the Apache License, Version 2.0 (the -->
 <!--- "License"); you may not use this file except in compliance -->
 <!--- with the License.  You may obtain a copy of the License at -->
 <!---   http://www.apache.org/licenses/LICENSE-2.0 -->
 <!--- Unless required by applicable law or agreed to in writing, -->
 <!--- software distributed under the License is distributed on an -->
 <!--- "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY -->
 <!--- KIND, either express or implied.  See the License for the -->
 <!--- specific language governing permissions and limitations -->
 <!--- under the License. -->
-Running PP-PicoDet via TVM on bare metal Arm(R) Cortex(R)-M55 CPU and CMSIS-NN
-===============================================================
-This folder contains an example of how to use TVM to run a PP-PicoDet model
+Running PP-PicoDet object detection model on bare metal Arm(R) Cortex(R)-M55 CPU using Arm Virtual Hardware
-on bare metal Cortex(R)-M55 CPU and CMSIS-NN.
+======================================================================
+This folder contains an example of how to run a PP-PicoDet model on bare metal [Cortex(R)-M55 CPU](https://www.arm.com/products/silicon-ip-cpu/cortex-m/cortex-m55) using [Arm Virtual Hardware](https://www.arm.com/products/development-tools/simulation/virtual-hardware).
-Prerequisites
+Running environment and prerequisites
 -------------
-If the demo is run in the ci_cpu Docker container provided with TVM, then the following
+Case 1: If the demo is run in Arm Virtual Hardware Amazon Machine Image(AMI) instance hosted by [AWS](https://aws.amazon.com/marketplace/pp/prodview-urbpq7yo5va7g?sr=0-1&ref_=beagle&applicationId=AWSMPContessa)/[AWS China](https://awsmarketplace.amazonaws.cn/marketplace/pp/prodview-2y7nefntbmybu), the following software will be installed through [configure_avh.sh](./configure_avh.sh) script. It will install automatically when you run the application through [run_demo.sh](./run_demo.sh) script.
-software will already be installed.
+You can refer to this [guide](https://arm-software.github.io/AVH/main/examples/html/MicroSpeech.html#amilaunch) to launch an Arm Virtual Hardware AMI instance.
-If the demo is not run in the ci_cpu Docker container, then you will need the following:
+Case 2: If the demo is run in the [ci_cpu Docker container](https://github.com/apache/tvm/blob/main/docker/Dockerfile.ci_cpu) provided with [TVM](https://github.com/apache/tvm), then the following software will already be installed.
+Case 3: If the demo is not run in the ci_cpu Docker container, then you will need the following:
 - Software required to build and run the demo (These can all be installed by running
  tvm/docker/install/ubuntu_install_ethosu_driver_stack.sh.)
  - [Fixed Virtual Platform (FVP) based on Arm(R) Corstone(TM)-300 software](https://release/2.5er.arm.com/tools-and-software/open-source-software/arm-platforms-software/arm-ecosystem-fvps)
@@ -37,26 +42,37 @@ If the demo is not run in the ci_cpu Docker container, then you will need the fo
    pip install -r ./requirements.txt
    ```
+In case2 and case3:  
+You will need to update your PATH environment variable to include the path to cmake 3.19.5 and the FVP.
+For example if you've installed these in ```/opt/arm``` , then you would do the following:
+```bash
+export PATH=/opt/arm/FVP_Corstone_SSE-300/models/Linux64_GCC-6.4:/opt/arm/cmake/bin:$PATH
+```
 You will also need TVM which can either be:
+  - Installed from TLCPack(see [TLCPack](https://tlcpack.ai/))
  - Built from source (see [Install from Source](https://tvm.apache.org/docs/install/from_source.html))
    - When building from source, the following need to be set in config.cmake:
      - set(USE_CMSISNN ON)
      - set(USE_MICRO ON)
      - set(USE_LLVM ON)
-  - Installed from TLCPack(see [TLCPack](https://tlcpack.ai/))
-You will need to update your PATH environment variable to include the path to cmake 3.19.5 and the FVP.
-For example if you've installed these in ```/opt/arm``` , then you would do the following:
-```bash
-export PATH=/opt/arm/FVP_Corstone_SSE-300/models/Linux64_GCC-6.4:/opt/arm/cmake/bin:$PATH
-```
 Running the demo application
 ----------------------------
 Type the following command to run the bare metal text recognition application ([src/demo_bare_metal.c](./src/demo_bare_metal.c)):
 ```bash
 ./run_demo.sh
 ```
+If you are not able to use Arm Virtual Hardware Amazon Machine Image(AMI) instance hosted by AWS/AWS China, specify argument --enable_FVP to 1 to make the application run on local Fixed Virtual Platforms (FVPs) executables.
+```bash
+./run_demo.sh --enable_FVP 1
+```
 If the Ethos(TM)-U platform and/or CMSIS have not been installed in /opt/arm/ethosu then
 the locations for these can be specified as arguments to run_demo.sh, for example:
@@ -65,13 +81,14 @@ the locations for these can be specified as arguments to run_demo.sh, for exampl
 --ethosu_platform_path /home/tvm-user/ethosu/core_platform
 ```
-This will:
+With [run_demo.sh](./run_demo.sh) to run the demo application, it will:
- Download a PP-PicoDet text recognition model
+- Set up running environment by installing the required prerequisites automatically if running in Arm Virtual Hardware Amazon AMI instance(not specify --enable_FVP to 1)
+- Download a PP-PicoDet model
 - Use tvmc to compile the text recognition model for Cortex(R)-M55 CPU and CMSIS-NN
 - Create a C header file inputs.c containing the image data as a C array
 - Create a C header file outputs.c containing a C array where the output of inference will be stored
 - Build the demo application
- Run the demo application on a Fixed Virtual Platform (FVP) based on Arm(R) Corstone(TM)-300 software
+- Run the demo application on a Arm Virtual Hardware based on Arm(R) Corstone(TM)-300 software
 - The application will report the text on the image and the corresponding score.
 Using your own image
@@ -82,9 +99,9 @@ image to be converted into an array of bytes for consumption by the model.
 The demo can be modified to use an image of your choice by changing the following line in run_demo.sh
 ```bash
-python3 ./convert_image.py ../../demo/000000014439_640x640.jpg
+python3 ./convert_image.py path/to/image
 ```
 Model description
 -----------------
-In this demo, the model we used is based on [PP-PicoDet](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.4/configs/picodet). Because of the excellent performance, PP-PicoDet are very suitable for deployment on mobile or CPU. And it is released by [PaddleDetection](https://github.com/PaddlePaddle/PaddleDetection).
+In this demo, the model we used is based on [PP-PicoDet](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.5/configs/picodet). Because of the excellent performance, PP-PicoDet are very suitable for deployment on mobile or CPU. And it is released by [PaddleDetection](https://github.com/PaddlePaddle/PaddleDetection).
--- a/deploy/third_engine/demo_avh/README.md.bak
+++ b/deploy/third_engine/demo_avh/README.md.bak
+<!--- Licensed to the Apache Software Foundation (ASF) under one -->
+<!--- or more contributor license agreements.  See the NOTICE file -->
+<!--- distributed with this work for additional information -->
+<!--- regarding copyright ownership.  The ASF licenses this file -->
+<!--- to you under the Apache License, Version 2.0 (the -->
+<!--- "License"); you may not use this file except in compliance -->
+<!--- with the License.  You may obtain a copy of the License at -->
+<!---   http://www.apache.org/licenses/LICENSE-2.0 -->
+<!--- Unless required by applicable law or agreed to in writing, -->
+<!--- software distributed under the License is distributed on an -->
+<!--- "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY -->
+<!--- KIND, either express or implied.  See the License for the -->
+<!--- specific language governing permissions and limitations -->
+<!--- under the License. -->
+Running PP-PicoDet via TVM on bare metal Arm(R) Cortex(R)-M55 CPU and CMSIS-NN
+===============================================================
+This folder contains an example of how to use TVM to run a PP-PicoDet model
+on bare metal Cortex(R)-M55 CPU and CMSIS-NN.
+Prerequisites
+-------------
+If the demo is run in the ci_cpu Docker container provided with TVM, then the following
+software will already be installed.
+If the demo is not run in the ci_cpu Docker container, then you will need the following:
+- Software required to build and run the demo (These can all be installed by running
+  tvm/docker/install/ubuntu_install_ethosu_driver_stack.sh.)
+  - [Fixed Virtual Platform (FVP) based on Arm(R) Corstone(TM)-300 software](https://developer.arm.com/tools-and-software/open-source-software/arm-platforms-software/arm-ecosystem-fvps)
+  - [cmake 3.19.5](https://github.com/Kitware/CMake/releases/)
+  - [GCC toolchain from Arm(R)](https://developer.arm.com/-/media/Files/downloads/gnu-rm/10-2020q4/gcc-arm-none-eabi-10-2020-q4-major-x86_64-linux.tar.bz2)
+  - [Arm(R) Ethos(TM)-U NPU driver stack](https://review.mlplatform.org)
+  - [CMSIS](https://github.com/ARM-software/CMSIS_5)
+- The python libraries listed in the requirements.txt of this directory
+  - These can be installed by running the following from the current directory:
+    ```bash
+    pip install -r ./requirements.txt
+    ```
+You will also need TVM which can either be:
+  - Built from source (see [Install from Source](https://tvm.apache.org/docs/install/from_source.html))
+    - When building from source, the following need to be set in config.cmake:
+      - set(USE_CMSISNN ON)
+      - set(USE_MICRO ON)
+      - set(USE_LLVM ON)
+  - Installed from TLCPack(see [TLCPack](https://tlcpack.ai/))
+You will need to update your PATH environment variable to include the path to cmake 3.19.5 and the FVP.
+For example if you've installed these in ```/opt/arm``` , then you would do the following:
+```bash
+export PATH=/opt/arm/FVP_Corstone_SSE-300/models/Linux64_GCC-6.4:/opt/arm/cmake/bin:$PATH
+```
+Running the demo application
+----------------------------
+Type the following command to run the bare metal text recognition application ([src/demo_bare_metal.c](./src/demo_bare_metal.c)):
+```bash
+./run_demo.sh
+```
+If the Ethos(TM)-U platform and/or CMSIS have not been installed in /opt/arm/ethosu then
+the locations for these can be specified as arguments to run_demo.sh, for example:
+```bash
+./run_demo.sh --cmsis_path /home/tvm-user/cmsis \
+--ethosu_platform_path /home/tvm-user/ethosu/core_platform
+```
+This will:
+- Download a PP-PicoDet text recognition model
+- Use tvmc to compile the text recognition model for Cortex(R)-M55 CPU and CMSIS-NN
+- Create a C header file inputs.c containing the image data as a C array
+- Create a C header file outputs.c containing a C array where the output of inference will be stored
+- Build the demo application
+- Run the demo application on a Fixed Virtual Platform (FVP) based on Arm(R) Corstone(TM)-300 software
+- The application will report the text on the image and the corresponding score.
+Using your own image
+--------------------
+The create_image.py script takes a single argument on the command line which is the path of the
+image to be converted into an array of bytes for consumption by the model.
+The demo can be modified to use an image of your choice by changing the following line in run_demo.sh
+```bash
+python3 ./convert_image.py ../../demo/000000014439_640x640.jpg
+```
+Model description
+-----------------
--- a/deploy/third_engine/demo_avh/configure_avh.sh
+++ b/deploy/third_engine/demo_avh/configure_avh.sh
+#!/bin/bash
+# Copyright (c) 2022 Arm Limited and Contributors. All rights reserved.
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+set -e
+set -u
+set -o pipefail
+# Show usage
+function show_usage() {
+    cat <<EOF
+Usage: Set up running environment by installing the required prerequisites.
+-h, --help
+    Display this help message.
+EOF
+}
+if [ "$#" -eq 1 ] && [ "$1" == "--help" -o "$1" == "-h" ]; then
+    show_usage
+    exit 0
+elif [ "$#" -ge 1 ]; then
+        show_usage
+        exit 1
+fi
+echo -e "\e[36mStart setting up running environment\e[0m"
+# Install CMSIS
+echo -e "\e[36mStart installing CMSIS\e[0m"
+CMSIS_PATH="/opt/arm/ethosu/cmsis" 
+mkdir -p "${CMSIS_PATH}"
+CMSIS_SHA="977abe9849781a2e788b02282986480ff4e25ea6"
+CMSIS_SHASUM="86c88d9341439fbb78664f11f3f25bc9fda3cd7de89359324019a4d87d169939eea85b7fdbfa6ad03aa428c6b515ef2f8cd52299ce1959a5444d4ac305f934cc"
+CMSIS_URL="http://github.com/ARM-software/CMSIS_5/archive/${CMSIS_SHA}.tar.gz"
+DOWNLOAD_PATH="/tmp/${CMSIS_SHA}.tar.gz"
+wget ${CMSIS_URL} -O "${DOWNLOAD_PATH}"
+echo "$CMSIS_SHASUM" ${DOWNLOAD_PATH} | sha512sum -c
+tar -xf "${DOWNLOAD_PATH}" -C "${CMSIS_PATH}" --strip-components=1
+touch "${CMSIS_PATH}"/"${CMSIS_SHA}".sha
+echo -e "\e[36mCMSIS Installation SUCCESS\e[0m"
+# Install Arm(R) Ethos(TM)-U NPU driver stack
+echo -e "\e[36mStart installing Arm(R) Ethos(TM)-U NPU driver stack\e[0m"
+git clone "https://review.mlplatform.org/ml/ethos-u/ethos-u-core-platform" /opt/arm/ethosu/core_platform 
+cd /opt/arm/ethosu/core_platform 
+git checkout tags/"21.11"
+echo -e "\e[36mArm(R) Ethos(TM)-U Core Platform Installation SUCCESS\e[0m"
+# Install Arm(R) GNU Toolchain
+echo -e "\e[36mStart installing Arm(R) GNU Toolchain\e[0m"
+mkdir -p /opt/arm/gcc-arm-none-eabi
+export gcc_arm_url='https://developer.arm.com/-/media/Files/downloads/gnu-rm/10-2020q4/gcc-arm-none-eabi-10-2020-q4-major-x86_64-linux.tar.bz2?revision=ca0cbf9c-9de2-491c-ac48-898b5bbc0443&la=en&hash=68760A8AE66026BCF99F05AC017A6A50C6FD832A'
+curl --retry 64 -sSL ${gcc_arm_url} | tar -C /opt/arm/gcc-arm-none-eabi --strip-components=1 -jx 
+export PATH=/opt/arm/gcc-arm-none-eabi/bin:$PATH 
+arm-none-eabi-gcc --version
+arm-none-eabi-g++ --version
+echo -e "\e[36mArm(R) Arm(R) GNU Toolchain Installation SUCCESS\e[0m"
+# Install TVM from TLCPack
+echo -e "\e[36mStart installing TVM\e[0m"
+pip install tlcpack-nightly -f https://tlcpack.ai/wheels
+echo -e "\e[36mTVM Installation SUCCESS\e[0m"
\ No newline at end of file
--- a/deploy/third_engine/demo_avh/convert_image.py
+++ b/deploy/third_engine/demo_avh/convert_image.py
@@ -24,10 +24,10 @@ import math
 from PIL import Image
 import numpy as np
 def resize_norm_img(img, image_shape, padding=True):
    imgC, imgH, imgW = image_shape
-    img = cv2.resize(
+    img = cv2.resize(img, (imgW, imgH), interpolation=cv2.INTER_LINEAR)
-        img, (imgW, imgH), interpolation=cv2.INTER_LINEAR)
    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
    img = np.transpose(img, [2, 0, 1]) / 255
    img = np.expand_dims(img, 0)
@@ -47,9 +47,8 @@ def create_header_file(name, tensor_name, tensor_data, output_path):
    raw_path = file_path.with_suffix(".h").resolve()
    with open(raw_path, "a") as header_file:
        header_file.write(
-            "\n"
+            "\n" + f"const size_t {tensor_name}_len = {tensor_data.size};\n" +
-            + f"const size_t {tensor_name}_len = {tensor_data.size};\n"
+            f'__attribute__((section(".data.tvm"), aligned(16))) float {tensor_name}[] = '
-            + f'__attribute__((section(".data.tvm"), aligned(16))) float {tensor_name}[] = '
        )
        header_file.write("{")
@@ -66,14 +65,16 @@ def create_headers(image_name):
    # Resize image to 32x320
    img = cv2.imread(img_path)
-    img = resize_norm_img(img, [3,32,320])
+    img = resize_norm_img(img, [3, 32, 320])
    img_data = img.astype("float32")
    # # Add the batch dimension, as we are expecting 4-dimensional input: NCHW.
    img_data = np.expand_dims(img_data, axis=0)
-    os.remove("./include/inputs.h")
+    if os.path.exists("./include/inputs.h"):
-    os.remove("./include/outputs.h")
+        os.remove("./include/inputs.h")
+    if os.path.exists("./include/outputs.h"):
+        os.remove("./include/outputs.h")
    # Create input header file
    create_header_file("inputs", "input", img_data, "./include")
    # Create output header file
@@ -82,15 +83,13 @@ def create_headers(image_name):
        "outputs",
        "output0",
        output_data,
-        "./include",
+        "./include", )
-    )
    output_data = np.zeros([170000], np.float)
    create_header_file(
        "outputs",
        "output1",
        output_data,
-        "./include",
+        "./include", )
-    )
 if __name__ == "__main__":

--- a/deploy/third_engine/demo_avh/corstone300.ld
+++ b/deploy/third_engine/demo_avh/corstone300.ld
@@ -247,10 +247,10 @@ SECTIONS
  } > DTCM
-  .bss.NoInit :
+  .bss.noinit (NOLOAD):
  {
    . = ALIGN(16);
-    *(.bss.NoInit)
+    *(.bss.noinit.*)
    . = ALIGN(16);
  } > DDR AT > DDR

--- a/deploy/third_engine/demo_avh/image/000000014439_640x640.jpg
+++ b/deploy/third_engine/demo_avh/image/000000014439_640x640.jpg
--- a/deploy/third_engine/demo_avh/run_demo.sh
+++ b/deploy/third_engine/demo_avh/run_demo.sh
@@ -15,7 +15,6 @@
 # KIND, either express or implied.  See the License for the
 # specific language governing permissions and limitations
 # under the License.
-export PATH=/opt/arm/FVP_Corstone_SSE-300/models/Linux64_GCC-6.4:/opt/arm/cmake/bin:$PATH
 set -e
 set -u
 set -o pipefail
@@ -25,18 +24,28 @@ function show_usage() {
    cat <<EOF
 Usage: run_demo.sh
 -h, --help
-    Display this help message.
+	Display this help message.
 --cmsis_path CMSIS_PATH
-    Set path to CMSIS.
+	Set path to CMSIS.
 --ethosu_platform_path ETHOSU_PLATFORM_PATH
-    Set path to Arm(R) Ethos(TM)-U core platform.
+	Set path to Arm(R) Ethos(TM)-U core platform.
 --fvp_path FVP_PATH
-   Set path to FVP.
+	Set path to FVP.
 --cmake_path
-   Set path to cmake.
+	Set path to cmake.
+--enable_FVP
+	Set 1 to run application on local Fixed Virtual Platforms (FVPs) executables.
 EOF
 }
+# Configure environment variables
+FVP_enable=0
+export PATH=/opt/arm/gcc-arm-none-eabi/bin:$PATH
+# Install python libraries
+echo -e "\e[36mInstall python libraries\e[0m"
+sudo pip install -r ./requirements.txt
 # Parse arguments
 while (( $# )); do
    case "$1" in
@@ -92,6 +101,18 @@ while (( $# )); do
                exit 1
            fi
            ;;
+        --enable_FVP)
+            if [ $# -gt 1 ] && [ "$2" == "1" -o "$2" == "0" ];
+            then
+                FVP_enable="$2"
+                shift 2
+            else
+                echo 'ERROR: --enable_FVP requires a right argument 1 or 0' >&2
+                show_usage >&2
+                exit 1
+            fi
+            ;;
        -*|--*)
            echo "Error: Unknown flag: $1" >&2
@@ -101,6 +122,16 @@ while (( $# )); do
    esac
 done
+# Choose running environment: cloud(default) or local environment
+Platform="VHT_Corstone_SSE-300_Ethos-U55"
+if [ $FVP_enable == "1" ]; then
+	Platform="FVP_Corstone_SSE-300_Ethos-U55"
+	echo -e "\e[36mRun application on local Fixed Virtual Platforms (FVPs)\e[0m"
+else
+	if [ ! -d "/opt/arm/" ]; then
+		sudo ./configure_avh.sh
+	fi
+fi
 # Directories
 script_dir="$( cd "$( dirname "${BASH_SOURCE[0]}" )" &> /dev/null && pwd )"
@@ -110,6 +141,11 @@ make cleanall
 mkdir -p build
 cd build
+# Get PaddlePaddle inference model
+echo -e "\e[36mDownload PaddlePaddle inference model\e[0m"
+wget https://bj.bcebos.com/v1/paddledet/deploy/Inference/picodet_s_320_coco_lcnet_no_nms.tar
+tar -xf picodet_s_320_coco_lcnet_no_nms.tar
 # Compile model for Arm(R) Cortex(R)-M55 CPU and CMSIS-NN
 # An alternative to using "python3 -m tvm.driver.tvmc" is to call
 # "tvmc" directly once TVM has been pip installed.
@@ -123,7 +159,7 @@ python3 -m tvm.driver.tvmc compile --target=cmsis-nn,c \
    --pass-config tir.usmp.enable=1 \
    --pass-config tir.usmp.algorithm=hill_climb \
    --pass-config tir.disable_storage_rewrite=1 \
-    --pass-config tir.disable_vectorize=1 ../models/picodet_s_320_coco_lcnet_no_nms/model \
+    --pass-config tir.disable_vectorize=1 picodet_s_320_coco_lcnet_no_nms/model.pdmodel \
    --output-format=mlf \
    --model-format=paddle \
    --module-name=picodet \
@@ -131,21 +167,18 @@ python3 -m tvm.driver.tvmc compile --target=cmsis-nn,c \
    --output=picodet.tar
 tar -xf picodet.tar
 # Create C header files
 cd ..
-python3 ./convert_image.py ../../demo/000000014439_640x640.jpg
+python3 ./convert_image.py ./image/000000014439_640x640.jpg
 # Build demo executable
-echo "Build demo executable..."
 cd ${script_dir}
 echo ${script_dir}
 make
-echo "End build demo executable..."
-# Run demo executable on the FVP
+# Run demo executable on the AVH
-FVP_Corstone_SSE-300_Ethos-U55 -C cpu0.CFGDTCMSZ=15 \
+$Platform -C cpu0.CFGDTCMSZ=15 \
 -C cpu0.CFGITCMSZ=15 -C mps3_board.uart0.out_file=\"-\" -C mps3_board.uart0.shutdown_tag=\"EXITTHESIM\" \
 -C mps3_board.visualisation.disable-visualisation=1 -C mps3_board.telnetterminal0.start_telnet=0 \
 -C mps3_board.telnetterminal1.start_telnet=0 -C mps3_board.telnetterminal2.start_telnet=0 -C mps3_board.telnetterminal5.start_telnet=0 \
-./build/demo
+./build/demo --stat