Skip to main content

Image Classification - MobileNetV2

This example demonstrates how to perform image classification using a BPU-deployed MobileNetV2 model with C/C++ inference. The example code is located in the directory /app/cdev_demo/bpu/01_classification_sample/02_mobilenetv2/.

Model Description​

  • Overview:

    MobileNetV2 is a lightweight convolutional neural network introduced by Google in 2018, designed for efficient image recognition on mobile devices. It incorporates the Inverted Residual and Linear Bottleneck structures to reduce computational cost while improving performance. MobileNetV2 is highly suitable for deployment on edge devices and in resource-constrained scenarios for tasks such as image classification and detection. The MobileNetV2 model used in this example accepts 224×224 input and is a BPU quantized model supporting the NV12 format.

  • HBM Model Name: mobilenetv2_224x224_nv12.hbm

  • Input Format: NV12, size 224x224 (separated Y and UV planes)

  • Output: Softmax probability distribution over 1000 classes (conforming to the ImageNet 1000-class standard)

Functionality Description​

  • Model Loading

    Load the model file and extract information such as model name, number of inputs and outputs, etc.

  • Input Preprocessing

    Resize the input image from BGR format to 224x224, then convert it to the hardware-required NV12 format (with Y and UV separated), and structure it into a dictionary for inference interface compatibility.

  • Inference Execution

    Call the .infer() method to perform inference.

  • Result Post-processing

    Retrieve the model output tensor, parse the probabilities, and display the top-K (default top-5) prediction results, including corresponding class names and probabilities.

Environment Dependencies​

Before compiling and running, ensure the following dependencies are installed:

sudo apt update
sudo apt install libgflags-dev

Directory Structure​

.
├── CMakeLists.txt              # CMake build script
├── README.md                   # Usage instructions
├── inc
│   └── mobilenetv2.hpp         # Model inference header file
└── src
    ├── main.cc                 # Main program entry point
    └── mobilenetv2.cc          # Model implementation

Build the Project​

  • Build the project 
    mkdir build && cd build
    cmake ..
    make -j$(nproc)

Model Download​

If the model is not found during program execution, download it using the following command:

wget https://archive.d-robotics.cc/downloads/rdk_model_zoo/rdk_s100/MobileNet/mobilenetv2_224x224_nv12.hbm

Parameter Description​

ParameterDescriptionDefault Value
--model_pathPath to the model file (.hbm format)/opt/hobot/model/s100/basic/mobilenetv2_224x224_nv12.hbm
--test_imgPath to the test image/app/res/assets/zebra_cls.jpg
--label_filePath to the class label mapping file (dict format)/app/res/labels/imagenet1000_clsidx_to_labels.txt
--top_kNumber of top-k classification results to output5

Quick Start​

  • Run the model
    • Ensure you are in the build directory.
    • Run with default parameters: 
      ./mobilenetv2
    • Run with custom parameters: 
      ./mobilenetv2 \
      --model_path /opt/hobot/model/s100/basic/mobilenetv2_224x224_nv12.hbm \
      --test_img /app/res/assets/zebra_cls.jpg \
      --label_file /app/res/labels/imagenet1000_clsidx_to_labels.txt \
      --top_k 5
  • View results 
    TOP 0: label=zebra, prob=0.992246
    TOP 1: label=tiger, Panthera tigris, prob=0.00404656
    TOP 2: label=hartebeest, prob=0.00133707
    TOP 3: label=tiger cat, prob=0.000722661
    TOP 4: label=impala, Aepyceros melampus, prob=0.000539704

Notes​

  • The output displays the top-K classes with the highest probabilities.

  • For more information on deployment methods or model support, please refer to the official documentation or contact platform technical support.

Last updated on