Skip to main content

Example Overview

This project includes multiple AI example programs written in C/C++ for the RDK S100 platform, covering common AI tasks such as image classification, object detection, instance segmentation, pose estimation, OCR, and speech recognition. The examples run inference with quantized models in .hbm format so developers can quickly validate model performance and start application development.

This project includes multiple AI example programs written in C/C++ for the RDK S600 platform, covering common AI tasks such as image classification, object detection, instance segmentation, pose estimation, and speech recognition. The examples run inference with quantized models in .hbm format so developers can quickly validate model performance and start application development.

On-device code for this project is located at: /app/cdev_demo/bpu.

Directory Structure Overview​

|-- 01_classification_sample         # Image classification examples (e.g. ResNet18, MobileNet)
|-- 02_detection_sample              # Object detection examples (e.g. YOLO)
|-- 03_instance_segmentation_sample  # Instance segmentation examples
|-- 04_pose_sample                   # Keypoint detection examples
|-- 05_open_instance_seg_sample      # Open-vocabulary instance segmentation examples
|-- 06_lane_detection_sample         # Lane detection examples
|-- 07_speech_sample                 # Speech recognition examples
|-- 08_OCR_sample                    # OCR examples
|-- 09_usb_camera_sample             # USB camera real-time inference examples
|-- 10_mipi_camera_sample            # MIPI camera real-time inference examples
|-- 11_decode_yolov5x_display_sample # Video decode, inference, and display examples
|-- 12_rtsp_yolov5x_display_sample   # RTSP stream decode, inference, and display examples
|-- utils                            # Shared utility functions
`-- README.md                        # Project documentation (this file)
|-- classification_sample            # Image classification examples (e.g. ResNet18, MobileNet)
|-- detection_sample                 # Object detection examples (e.g. YOLO)
|-- instance_segmentation_sample     # Instance segmentation examples
|-- pose_sample                      # Keypoint detection examples
|-- speech_sample                    # Speech recognition examples
|-- usb_camera_sample                # USB camera real-time inference examples
|-- mipi_camera_sample               # MIPI camera real-time inference examples
|-- decode_yolov5x_display_sample    # Video decode, inference, and display examples
|-- rtsp_yolov5x_display_sample      # RTSP stream decode, inference, and display examples
|-- utils                            # Shared utility functions
`-- README.md                        # Project documentation (this file)

Environment Requirements​

Before running the examples, ensure your system meets the following requirements:

Hardware​

  • RDK S100 development board with BPU support
  • Camera (USB or MIPI) if you plan to run camera-related examples
  • RDK S600 development board with BPU support
  • Camera (USB or MIPI) if you plan to run camera-related examples

System and Toolchain​

This project has been verified to run in the following environment:

  • Operating system

    • Ubuntu 22.04.5 LTS (Jammy Jellyfish)
    • Distributor ID: Ubuntu
    • Description: Ubuntu 24.04.3 LTS
    • Release: 24.04
    • Codename: noble

  • Build toolchain

    • CMake: 3.22.1

    • GCC: 11.4.0

    • G++: 11.4.0

Dependencies​

Different examples require different development packages. Install them as needed.

  • Common dependencies

    sudo apt update
    sudo apt install libgflags-dev

  • ASR (speech recognition) example

    sudo apt update
    sudo apt install libsndfile1-dev
    sudo apt install libsamplerate0-dev
  • OCR example 
    sudo apt update
    sudo apt install libpolyclipping-dev

Build Instructions​

Using the ResNet18 image classification example:

cd 01_classification_sample/01_resnet18

mkdir build && cd build

cmake ..

make -j$(nproc)
cd classification_sample/resnet18

mkdir build && cd build

cmake ..

make -j$(nproc)

Running Examples​

Using the ResNet18 image classification example:

  • Go to the sample build directory

    cd 01_classification_sample/01_resnet18/build
    cd classification_sample/resnet18/build

  • Run the model

    ./resnet_18

  • View results

    TOP 0: label=zebra, prob=0.99872
    TOP 1: label=cheetah, chetah, Acinonyx jubatus, prob=0.000448407
    TOP 2: label=impala, Aepyceros melampus, prob=0.000398787
    TOP 3: label=gazelle, prob=0.000253181
    TOP 4: label=prairie chicken, prairie grouse, prairie fowl, prob=0.000179423

Common Utilities​

The utils directory contains shared utility functions used across BPU C/C++ inference examples, including image preprocessing, inference post-processing, multimedia processing, and general helpers for reuse across samples.


utils
├── inc                          # Header files
│   ├── common_utils.hpp         # General utilities (dequantization, result drawing, common data structures, etc.)
│   ├── multimedia_utils.hpp     # Multimedia utilities (video frame decode, pixel format conversion, etc.)
│   ├── postprocess_utils.hpp    # Inference post-processing (NMS, decoding, mask handling, etc.)
│   └── preprocess_utils.hpp     # Input preprocessing (image resize, normalization, format conversion, etc.)
└── src                          # Source implementations
    ├── common_utils.cc
    ├── multimedia_utils.cc
    ├── postprocess_utils.cc
    └── preprocess_utils.cc


Additional Notes​

  • All examples use models in .hbm format.

  • Each subdirectory includes a README.md with environment requirements, command-line arguments, and run instructions for that model.

Last updated on