Model Zoo
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 unless otherwise explicitly stated.