Skip to main content

4.1 Demo Zoo Overview

Intro

The Bianbu AI Demo Zoo is a collection of sample projects independently developed by SpacemiT. It is designed to provide deployment references for a variety of deep learning models on the K1 series chips, demonstrating an end-to-end inference workflow.

This project offers two main branches: Computer Vision (CV) and Natural Language Processing (NLP). The CV branch covers typical tasks such as image classification, object detection, and face recognition. It supports both C++ and Python development languages and is suitable for a wide range of real-world deployment scenariosa.

Project address:⭐ Bianbu AI Demo Zoo

For a detailed list of supported models, please see the Model List.Currently, it covers common models such as classification networks (ResNet, MobileNet), detection networks (YOLOv5, YOLOX), and face recognition models (ArcFace).

Demos

Most models provide inference samples in both Python and C++. After downloading the necessary model weights and test data according to the README.md file, users can complete the inference demonstration in just a few steps.

Python Demo

Taking ResNet image classification as an example:

cd python
python test_resnet.py # Default to using the ResNet50 model.

After the model finishes running, it will output the predicted class labels.

C++ Demo

cd cpp
mkdir build
cd build
cmake ..
make -j8
./resnet_demo --model /path/to/resnet50.onnx --image /path/to/image.jpg

After it's done running, the prediction results will also be output in the terminal. You can combine this with OpenCV to render the classification labels.