README.md

---
license: mit
language:
- en
metrics:
- accuracy
pipeline_tag: image-classification
tags:
- biology
---

# Bird-Species-Classification

This project only to show a demo of Bird Species Classification model within 1400+ species of birds.

The model is trained with 224x224 resolution.

This model has been converted to run on the Axera NPU using **w8a16** quantization.

This model has been optimized with the following LoRA: 

Compatible with Pulsar2 version: 5.1

## Convert tools links:

For those who are interested in model conversion, you can try to export axmodel through 

- [Pulsar2 Link, How to Convert ONNX to axmodel](https://pulsar2-docs.readthedocs.io/en/latest/pulsar2/introduction.html) 


## Support Platform

- AX650
  - [M4N-Dock(爱芯派Pro)](https://wiki.sipeed.com/hardware/zh/maixIV/m4ndock/m4ndock.html)
  - [M.2 Accelerator card](https://docs.m5stack.com/zh_CN/ai_hardware/LLM-8850_Card)
- AX630C
  - [爱芯派2](https://axera-pi-2-docs-cn.readthedocs.io/zh-cn/latest/index.html)
  - [Module-LLM](https://docs.m5stack.com/zh_CN/module/Module-LLM)
  - [LLM630 Compute Kit](https://docs.m5stack.com/zh_CN/core/LLM630%20Compute%20Kit)
 
| Models       | Platforms    | latency       | Top5 Accuracy | CMM size(MB)  |
| -------------| -------------| ------------- | --------------| --------------|
|              | AX650        | 0.19ms        |               |               |
| bird-s       | AX630C       | 0.54ms        | 66%           | 1.07          |
|              | AX615        | 0.87ms        |               |               |
|              | AX650        | 0.58ms        |               |               |
| bird-m       | AX630C       | 2.52ms        | 79%           | 12.2          |
|              | AX615        | 4.97ms        |               |               |

## How to use

Download all files from this repository to the device

```
root@ax650:~/Bird-Species-Classification# tree
.
├── README.md
├── axmodel_infer.py
├── class_name.txt
├── model
│   ├── AX615
│   │   ├── bird_615_npu1.axmodel
│   │   └── bird_615_npu2.axmodel
│   ├── AX620E
│   │   ├── bird_630_npu1.axmodel
│   │   └── bird_630_npu2.axmodel
│   └── AX650
│       └── bird_650_npu3.axmodel
├── onnx_infer.py
├── prediction_result_top5.png
├── quant
│   ├── Bird.json
│   ├── README.md
│   └── bird.tar.gz
└── test_images
    ├── 03111_2c0dfa5a-c4a0-47f8-ac89-6a289208050f.jpg
    ├── 03332_01b365c3-a741-4f45-bac2-4345bc901ec6.jpg
    ├── 03412_0ffc115b-43b4-4474-a373-24233f391de3.jpg
    ├── 03615_0dfbf6ae-434d-4648-b5d2-08412546ea64.jpg
    ├── 04251_3a52191e-be71-4539-98ea-14a8f2347330.jpg
    ├── 04405_0c5a6785-0bc2-49d9-9702-b9e94ba9b686.jpg
    └── 04593_3d74d5a7-15b1-4bb9-af6f-1bcd78485787.jpg

6 directories, 20 files

```

### python env requirement

#### pyaxengine

https://github.com/AXERA-TECH/pyaxengine

```
wget https://github.com/AXERA-TECH/pyaxengine/releases/download/0.1.3rc0/axengine-0.1.3-py3-none-any.whl
pip install axengine-0.1.3-py3-none-any.whl
```

## Inference with AX650 Host, such as M4N-Dock(爱芯派Pro)

```
root@ax650:~/Bird-Species-Classification# python3 axmodel_infer.py --image test_images/04251_3a52191e-be71-4539-98ea-14a8f2347330.jpg
[INFO] Available providers:  ['AxEngineExecutionProvider']
Loading ONNX model with providers: ['AxEngineExecutionProvider']
[INFO] Using provider: AxEngineExecutionProvider
[INFO] Chip type: ChipType.MC50
[INFO] VNPU type: VNPUType.DISABLED
[INFO] Engine version: 2.12.0s
[INFO] Model type: 2 (triple core)
[INFO] Compiler version: 5.1-patch1 74996179

Image: test_images/04251_3a52191e-be71-4539-98ea-14a8f2347330.jpg
Top-5 Predictions:
#1: 04251_Animalia_Chordata_Aves_Passeriformes_Tityridae_Tityra_semifasciata (0.9823)
#2: 04333_Animalia_Chordata_Aves_Passeriformes_Tyrannidae_Tyrannus_savana (0.0103)
#3: 04208_Animalia_Chordata_Aves_Passeriformes_Sylviidae_Sylvia_melanocephala (0.0056)
#4: 03561_Animalia_Chordata_Aves_Coraciiformes_Alcedinidae_Todiramphus_chloris (0.0009)
#5: 04455_Animalia_Chordata_Aves_Piciformes_Picidae_Melanerpes_lewis (0.0005)
Result saved to: prediction_result_top5.png
```

output：
<img src="./prediction_result_top5.png" width="200%" alt="预测结果" />