Hugging Face's logo

STMicroelectronics
/
GMP-HAR-model

Model card Files
xet
Community
FBAGSTM commited on
Commit
89c2163
·
verified ·
1 Parent(s): 3a409a9

Update Readme ST Model Zoo

Browse files
Files changed (1) hide show
  1. README.md +10 -13
README.md CHANGED
@@ -1,9 +1,3 @@
1
- ---
2
- license: other
3
- license_name: sla0044
4
- license_link: >-
5
- https://github.com/STMicroelectronics/stm32ai-modelzoo/tree/main/human_activity_recognition/gmp/ST_pretrainedmodel_public_dataset/LICENSE.md
6
- ---
7
  # GMP HAR model
8
 
9
  ## **Use case** : `Human activity recognition`
@@ -54,7 +48,7 @@ For a frame of resolution of (wl x 3) and P classes
54
 
55
  ## Metrics
56
 
57
- Measures are done with default STM32Cube.AI Dev Cloud version 10.0.0 and for target board B-U585I-IOT02A. In addition the configuration were enabled input / output allocated option and `balanced` as optimization choice.
58
 
59
  The inference time is reported is calculated on STM32 board **B-U585I-IOT02A** running at Frequency of **160 MHz**.
60
 
@@ -64,8 +58,8 @@ The inference time is reported is calculated on STM32 board **B-U585I-IOT02A** r
64
 
65
  | Model | Format | Input Shape | Target Board | Activation RAM (KiB) | Runtime RAM (KiB) | Weights Flash (KiB) | Code Flash (KiB) | Total RAM (KiB) | Total Flash (KiB) | Inference Time (ms) | STM32Cube.AI version |
66
  |:----------------------------------------------------------------------------:|:------:|:-----------:|:-------:|:--------------------:|:-----------------:|:-------------------:|:----------------:|:-----------------:|:-----------------:|:---------------------:|:---------------------:|
67
- | [GMP wl 24](https://github.com/STMicroelectronics/stm32ai-modelzoo/tree/main/human_activity_recognition/gmp/ST_pretrainedmodel_public_dataset/WISDM/gmp_wl_24/gmp_wl_24.h5) | FLOAT32| 24 x 3 x 1 | B-U585I-IOT02A | 4.25 | 2.08 | 5.70 | 12.29 | 6.33 | 18.96 | 4.42 | 10.0.0 |
68
- | [GMP wl 48](https://github.com/STMicroelectronics/stm32ai-modelzoo/tree/main/human_activity_recognition/gmp/ST_pretrainedmodel_public_dataset/WISDM/gmp_wl_48/gmp_wl_48.h5) | FLOAT32| 48 x 3 x 1 | B-U585I-IOT02A | 8.83 | 2.08 | 5.70 | 12.29 | 10.91 | 18.96 | 10.64 | 10.0.0 |
69
 
70
 
71
 
@@ -77,10 +71,14 @@ Dataset details: A custom dataset and not publically available, Number of classe
77
 
78
  | Model | Format | Resolution | Accuracy (%) |
79
  |:----------------------------------------------------------------------------------------------:|:--------:|:----------:|:-------------:|
80
- | [GMP wl 24](https://github.com/STMicroelectronics/stm32ai-modelzoo/tree/main/human_activity_recognition/gmp/ST_pretrainedmodel_custom_dataset/mobility_v1/gmp_wl_24/gmp_wl_24.h5) | FLOAT32 | 24 x 3 x 1 | 94.08 |
81
- | [GMP wl 48](https://github.com/STMicroelectronics/stm32ai-modelzoo/tree/main/human_activity_recognition/gmp/ST_pretrainedmodel_custom_dataset/mobility_v1/gmp_wl_48/gmp_wl_48.h5) | FLOAT32 | 48 x 3 x 1 | 93.84 |
82
 
83
 
 
 
 
 
84
  ### Accuracy with WISDM dataset
85
 
86
 
@@ -101,5 +99,4 @@ Please refer to the stm32ai-modelzoo-services GitHub [here](https://github.com/S
101
  # References
102
 
103
  <a id="1">[1]</a>
104
- “WISDM : Human activity recognition datasets". [Online]. Available: "https://www.cis.fordham.edu/wisdm/dataset.php".
105
-
 
 
 
 
 
 
 
1
  # GMP HAR model
2
 
3
  ## **Use case** : `Human activity recognition`
 
48
 
49
  ## Metrics
50
 
51
+ Measures are done with default STM32Cube.AI Dev Cloud version 10.2.0 and for target board B-U585I-IOT02A. In addition the configuration were enabled input / output allocated option and `balanced` as optimization choice.
52
 
53
  The inference time is reported is calculated on STM32 board **B-U585I-IOT02A** running at Frequency of **160 MHz**.
54
 
 
58
 
59
  | Model | Format | Input Shape | Target Board | Activation RAM (KiB) | Runtime RAM (KiB) | Weights Flash (KiB) | Code Flash (KiB) | Total RAM (KiB) | Total Flash (KiB) | Inference Time (ms) | STM32Cube.AI version |
60
  |:----------------------------------------------------------------------------:|:------:|:-----------:|:-------:|:--------------------:|:-----------------:|:-------------------:|:----------------:|:-----------------:|:-----------------:|:---------------------:|:---------------------:|
61
+ | [GMP wl 24](ST_pretrainedmodel_public_dataset/WISDM/gmp_wl_24/gmp_wl_24.h5) | FLOAT32| 24 x 3 x 1 | B-U585I-IOT02A | 4.25 | 2.08 | 5.70 | 12.29 | 6.33 | 18.96 | 4.42 | 10.2.0 |
62
+ | [GMP wl 48](ST_pretrainedmodel_public_dataset/WISDM/gmp_wl_48/gmp_wl_48.h5) | FLOAT32| 48 x 3 x 1 | B-U585I-IOT02A | 8.83 | 2.08 | 5.70 | 12.29 | 10.91 | 18.96 | 10.64 | 10.2.0 |
63
 
64
 
65
 
 
71
 
72
  | Model | Format | Resolution | Accuracy (%) |
73
  |:----------------------------------------------------------------------------------------------:|:--------:|:----------:|:-------------:|
74
+ | [GMP wl 24](./ST_pretrainedmodel_custom_dataset/mobility_v1/gmp_wl_24/gmp_wl_24.h5) | FLOAT32 | 24 x 3 x 1 | 94.08 |
75
+ | [GMP wl 48](./ST_pretrainedmodel_custom_dataset/mobility_v1/gmp_wl_48/gmp_wl_48.h5) | FLOAT32 | 48 x 3 x 1 | 93.84 |
76
 
77
 
78
+ Confusion matrix for GMP wl 24 with Float32 weights for mobility_v1 dataset is given below.
79
+
80
+ ![plot](./doc/img/mobility_v1_gmp_wl_24_confusion_matrix.png)
81
+
82
  ### Accuracy with WISDM dataset
83
 
84
 
 
99
  # References
100
 
101
  <a id="1">[1]</a>
102
+ “WISDM : Human activity recognition datasets". [Online]. Available: "https://www.cis.fordham.edu/wisdm/dataset.php".