object_classification

Running PaddleClas image classification model using Arm Virtual Hardware

Running environment and prerequisites

If you are running this project for the first time, please go back to the homepage and follow the documentation there to install the dependencies.

Supported Models List

For your convenience, we have provided a corresponding table of models currently adapted and the devices that support each of these models.

Model	Cortex-M55	Cortex-M85
MobileNetV1	✅	✅
MobileNetV3	✅	✅
PP_LCNet	✅	✅

Run Classification Model Demo

We use run_demo.sh to build the entire project. The parameters you need to pass are listed in the table below.

Arguments	Details
--model	The '--model' parameter is used to select the desired model. Currently, it supports MobileNetV1/MobileNetV3/PP_LCNet.
--device	The '--device' parameter is used to select the desired device. Currently, it supports cortex-m55/cortex-m85.

For example, running the following command can execute the MobileNetV3 model on Cortex-M55:

cd /path/to/Paddle-examples-for-AVH/object_classification
bash run_demo.sh --model MobileNetV3 --device cortex-m55