https://github.com/ARM-software/ML-zoo

https://github.com/ARM-software/ML-zoo

Last synced: 3 months ago
JSON representation

• Host: GitHub
• URL: https://github.com/ARM-software/ML-zoo
• Owner: ARM-software
• License: apache-2.0
• Created: 2020-10-14T16:17:33.000Z (over 4 years ago)
• Default Branch: master
• Last Pushed: 2023-03-31T09:47:56.000Z (almost 2 years ago)
• Last Synced: 2024-08-07T13:27:51.429Z (6 months ago)
• Language: Python
• Size: 61.4 MB
• Stars: 187
• Watchers: 12
• Forks: 51
• Open Issues: 9
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-keyword-spotting - Github: Micronets: Neural network architectures for deploying tinyml applications on commodity microcontrollers

README

        
# Model Zoo

![version](https://img.shields.io/badge/version-21.08-0091BD)

> A collection of machine learning models optimized for Arm IP.

## Anomaly Detection

    

        Network

        Type

        Framework

        Cortex-A

        Cortex-M

        Mali GPU

        Ethos U

        Score (AUC)

    

    

        MicroNet Large INT8 

        INT8

        TensorFlow Lite

        :heavy_multiplication_x: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.968

    

    

        MicroNet Medium INT8 

        INT8

        TensorFlow Lite

        :heavy_multiplication_x: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.963

    

    

        MicroNet Small INT8 

        INT8

        TensorFlow Lite

        :heavy_multiplication_x: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.955

    

**Dataset**: Dcase 2020 Task 2 Slide Rail

## Image Classification

    

        Network

        Type

        Framework

        Cortex-A

        Cortex-M

        Mali GPU

        Ethos U

        Score (Top 1 Accuracy)

    

    

        MobileNet v2 1.0 224 INT8 *

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.697

    

    

        MobileNet v2 1.0 224 UINT8 

        UINT8

        TensorFlow Lite

        :heavy_multiplication_x: 

        :heavy_multiplication_x: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.708

    

**Dataset**: ILSVRC 2012

## Keyword Spotting

    

        Network

        Type

        Framework

        Cortex-A

        Cortex-M

        Mali GPU

        Ethos U

        Score (Accuracy)

    

    

        CNN Large INT8 *

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.923

    

    

        CNN Medium INT8 *

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.905

    

    

        CNN Small INT8 *

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.902

    

    

        DNN Large INT8 *

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.860

    

    

        DNN Medium INT8 *

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.839

    

    

        DNN Small INT8 *

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.821

    

    

        DS-CNN Large Clustered FP32 *

        FP32

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        0.948

    

    

        DS-CNN Large Clustered INT8 *

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.939

    

    

        DS-CNN Large INT8 *

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: HERO

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.945

    

    

        DS-CNN Medium INT8 *

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: HERO

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.939

    

    

        DS-CNN Small INT8 *

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: HERO

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.931

    

        

        DS-CNN Small INT16 *

        INT16

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: HERO

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.934

    

    

        CNN Large FP32 *

        FP32

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        0.934

    

    

        CNN Medium FP32 *

        FP32

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        0.918

    

    

        CNN Small FP32 *

        FP32

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        0.922

    

    

        DNN Large FP32 *

        FP32

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        0.867

    

    

        DNN Medium FP32 *

        FP32

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        0.850

    

    

        DNN Small FP32 *

        FP32

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        0.836

    

    

        DS-CNN Large FP32 *

        FP32

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: HERO

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        0.950

    

    

        DS-CNN Medium FP32 *

        FP32

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: HERO

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        0.943

    

    

        DS-CNN Small FP32 *

        FP32

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: HERO

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        0.939

    

    

        MicroNet Large INT8 

        INT8

        TensorFlow Lite

        :heavy_multiplication_x: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.965

    

    

        MicroNet Medium INT8 

        INT8

        TensorFlow Lite

        :heavy_multiplication_x: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.958

    

    

        MicroNet Small INT8 

        INT8

        TensorFlow Lite

        :heavy_multiplication_x: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.953

    

**Dataset**: Google Speech Commands Test Set

## Noise Suppression

    

        Network

        Type

        Framework

        Cortex-A

        Cortex-M

        Mali GPU

        Ethos U

        Score (Average Pesq)

    

    

        RNNoise INT8 *

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        2.945

    

**Dataset**: Noisy Speech Database For Training Speech Enhancement Algorithms And Tts Models

## Object Detection

    

        Network

        Type

        Framework

        Cortex-A

        Cortex-M

        Mali GPU

        Ethos U

        Score (mAP)

    

    

        SSD MobileNet v1 FP32 *

        FP32

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        0.210

    

    

        SSD MobileNet v1 INT8 *

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        0.234

    

    

        SSD MobileNet v1 UINT8 *

        UINT8

        TensorFlow Lite

        :heavy_multiplication_x: 

        :heavy_multiplication_x: 

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        0.180

    

    

        YOLO v3 Tiny FP32 *

        FP32

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        0.331

    

**Dataset**: COCO Validation 2017

## Speech Recognition

    

        Network

        Type

        Framework

        Cortex-A

        Cortex-M

        Mali GPU

        Ethos U

        Score (LER)

    

    

        Wav2letter INT8 

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.0877

    

    

        Wav2letter Pruned INT8 *

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.0783

    

    

        Tiny Wav2letter INT8 *

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        :heavy_check_mark: 

        0.0348

    

    

        Tiny Wav2letter Pruned INT8 *

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        :heavy_check_mark: 

        0.0283

    

**Dataset**: LibriSpeech, Fluent Speech

## Superresolution

    

        Network

        Type

        Framework

        Cortex-A

        Cortex-M

        Mali GPU

        Ethos U

        Score (PSNR)

    

    

        SESR INT8 **

        INT8

        TensorFlow Lite

        :heavy_check_mark: 

        :heavy_multiplication_x: 

        :heavy_check_mark: HERO

        :heavy_multiplication_x: 

        35.00dB

    

**Dataset**: DIV2K

## Visual Wake Words

    

        Network

        Type

        Framework

        Cortex-A

        Cortex-M

        Mali GPU

        Ethos U

        Score (Accuracy)

    

    

        MicroNet VWW-2 INT8 

        INT8

        TensorFlow Lite

        :heavy_multiplication_x: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.768

    

    

        MicroNet VWW-3 INT8 

        INT8

        TensorFlow Lite

        :heavy_multiplication_x: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.855

    

    

        MicroNet VWW-4 INT8 

        INT8

        TensorFlow Lite

        :heavy_multiplication_x: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        :heavy_check_mark: 

        0.822

    

**Dataset**: Visual Wake Words

### Key

* :heavy_check_mark: - Will run on this platform.

* :heavy_multiplication_x: - Will not run on this platform.

* `*` - Code to recreate model available.

* `**` - This model has a large memory footprint – it will not run on all platforms.

## License

[Apache-2.0](https://spdx.org/licenses/Apache-2.0.html) unless otherwise explicitly stated.