Hugging Face's logo

qualcomm
/
FFNet-40S

Image Segmentation
PyTorch
real_time
android
Model card Files
xet
 Community
qaihm-bot commited on
Commit
b866df6
·
verified ·
1 Parent(s): 53e1ddc

v0.49.1

Browse files

See https://github.com/qualcomm/ai-hub-models/releases/v0.49.1 for changelog.

Files changed (1) hide show
  1. README.md +76 -76
README.md CHANGED
@@ -15,7 +15,7 @@ pipeline_tag: image-segmentation
15
  FFNet-40S is a "fuss-free network" that segments street scene images with per-pixel classes like road, sidewalk, and pedestrian. Trained on the Cityscapes dataset.
16
 
17
  This is based on the implementation of FFNet-40S found [here](https://github.com/Qualcomm-AI-research/FFNet).
18
- This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/ffnet_40s) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
19
 
20
  Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) to run these models on a hosted Qualcomm® device.
21
 
@@ -28,26 +28,26 @@ Below are pre-exported model assets ready for deployment.
28
 
29
  | Runtime | Precision | Chipset | SDK Versions | Download |
30
  |---|---|---|---|---|
31
- | ONNX | float | Universal | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/ffnet_40s/releases/v0.48.0/ffnet_40s-onnx-float.zip)
32
- | ONNX | w8a8 | Universal | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/ffnet_40s/releases/v0.48.0/ffnet_40s-onnx-w8a8.zip)
33
- | QNN_DLC | float | Universal | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/ffnet_40s/releases/v0.48.0/ffnet_40s-qnn_dlc-float.zip)
34
- | QNN_DLC | w8a8 | Universal | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/ffnet_40s/releases/v0.48.0/ffnet_40s-qnn_dlc-w8a8.zip)
35
- | TFLITE | float | Universal | QAIRT 2.43, TFLite 2.17.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/ffnet_40s/releases/v0.48.0/ffnet_40s-tflite-float.zip)
36
- | TFLITE | w8a8 | Universal | QAIRT 2.43, TFLite 2.17.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/ffnet_40s/releases/v0.48.0/ffnet_40s-tflite-w8a8.zip)
37
 
38
  For more device-specific assets and performance metrics, visit **[FFNet-40S on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/ffnet_40s)**.
39
 
40
 
41
  ### Option 2: Export with Custom Configurations
42
 
43
- Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/ffnet_40s) Python library to compile and export the model with your own:
44
  - Custom weights (e.g., fine-tuned checkpoints)
45
  - Custom input shapes
46
  - Target device and runtime configurations
47
 
48
  This option is ideal if you need to customize the model beyond the default configuration provided here.
49
 
50
- See our repository for [FFNet-40S on GitHub](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/ffnet_40s) for usage instructions.
51
 
52
  ## Model Details
53
 
@@ -64,73 +64,73 @@ See our repository for [FFNet-40S on GitHub](https://github.com/qualcomm/ai-hub-
64
  ## Performance Summary
65
  | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
66
  |---|---|---|---|---|---|---
67
- | FFNet-40S | ONNX | float | Snapdragon® X2 Elite | 13.263 ms | 22 - 22 MB | NPU
68
- | FFNet-40S | ONNX | float | Snapdragon® X Elite | 31.492 ms | 24 - 24 MB | NPU
69
- | FFNet-40S | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 22.483 ms | 30 - 303 MB | NPU
70
- | FFNet-40S | ONNX | float | Qualcomm® QCS8550 (Proxy) | 31.785 ms | 24 - 26 MB | NPU
71
- | FFNet-40S | ONNX | float | Qualcomm® QCS9075 | 48.061 ms | 24 - 27 MB | NPU
72
- | FFNet-40S | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 16.024 ms | 8 - 209 MB | NPU
73
- | FFNet-40S | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 12.347 ms | 26 - 247 MB | NPU
74
- | FFNet-40S | ONNX | w8a8 | Snapdragon® X2 Elite | 6.929 ms | 9 - 9 MB | NPU
75
- | FFNet-40S | ONNX | w8a8 | Snapdragon® X Elite | 10.461 ms | 8 - 8 MB | NPU
76
- | FFNet-40S | ONNX | w8a8 | Snapdragon® 8 Gen 3 Mobile | 6.568 ms | 7 - 248 MB | NPU
77
- | FFNet-40S | ONNX | w8a8 | Qualcomm® QCS6490 | 360.076 ms | 202 - 239 MB | CPU
78
- | FFNet-40S | ONNX | w8a8 | Qualcomm® QCS8550 (Proxy) | 9.833 ms | 0 - 14 MB | NPU
79
- | FFNet-40S | ONNX | w8a8 | Qualcomm® QCS9075 | 12.946 ms | 6 - 9 MB | NPU
80
- | FFNet-40S | ONNX | w8a8 | Qualcomm® QCM6690 | 365.578 ms | 215 - 224 MB | CPU
81
- | FFNet-40S | ONNX | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 7.726 ms | 1 - 192 MB | NPU
82
- | FFNet-40S | ONNX | w8a8 | Snapdragon® 7 Gen 4 Mobile | 361.936 ms | 164 - 173 MB | CPU
83
- | FFNet-40S | ONNX | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 6.563 ms | 2 - 198 MB | NPU
84
- | FFNet-40S | QNN_DLC | float | Snapdragon® X2 Elite | 14.557 ms | 24 - 24 MB | NPU
85
- | FFNet-40S | QNN_DLC | float | Snapdragon® X Elite | 37.107 ms | 24 - 24 MB | NPU
86
- | FFNet-40S | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 24.737 ms | 24 - 303 MB | NPU
87
- | FFNet-40S | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 134.465 ms | 15 - 210 MB | NPU
88
- | FFNet-40S | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 36.181 ms | 24 - 200 MB | NPU
89
- | FFNet-40S | QNN_DLC | float | Qualcomm® SA8775P | 48.805 ms | 24 - 220 MB | NPU
90
- | FFNet-40S | QNN_DLC | float | Qualcomm® QCS9075 | 62.122 ms | 24 - 52 MB | NPU
91
- | FFNet-40S | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 116.564 ms | 23 - 307 MB | NPU
92
- | FFNet-40S | QNN_DLC | float | Qualcomm® SA7255P | 134.465 ms | 15 - 210 MB | NPU
93
- | FFNet-40S | QNN_DLC | float | Qualcomm® SA8295P | 64.889 ms | 24 - 226 MB | NPU
94
- | FFNet-40S | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 18.149 ms | 21 - 248 MB | NPU
95
- | FFNet-40S | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 12.576 ms | 24 - 262 MB | NPU
96
- | FFNet-40S | QNN_DLC | w8a8 | Snapdragon® X2 Elite | 6.227 ms | 6 - 6 MB | NPU
97
- | FFNet-40S | QNN_DLC | w8a8 | Snapdragon® X Elite | 15.979 ms | 6 - 6 MB | NPU
98
- | FFNet-40S | QNN_DLC | w8a8 | Snapdragon® 8 Gen 3 Mobile | 10.571 ms | 6 - 249 MB | NPU
99
- | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® QCS6490 | 67.566 ms | 4 - 12 MB | NPU
100
- | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® QCS8275 (Proxy) | 32.881 ms | 6 - 198 MB | NPU
101
- | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® QCS8550 (Proxy) | 15.145 ms | 6 - 8 MB | NPU
102
- | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® SA8775P | 15.757 ms | 6 - 199 MB | NPU
103
- | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® QCS9075 | 18.22 ms | 6 - 14 MB | NPU
104
- | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® QCM6690 | 127.392 ms | 6 - 239 MB | NPU
105
- | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® QCS8450 (Proxy) | 21.509 ms | 6 - 250 MB | NPU
106
- | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® SA7255P | 32.881 ms | 6 - 198 MB | NPU
107
- | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® SA8295P | 20.179 ms | 6 - 201 MB | NPU
108
- | FFNet-40S | QNN_DLC | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 7.174 ms | 6 - 213 MB | NPU
109
- | FFNet-40S | QNN_DLC | w8a8 | Snapdragon® 7 Gen 4 Mobile | 18.969 ms | 6 - 220 MB | NPU
110
- | FFNet-40S | QNN_DLC | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 5.211 ms | 6 - 236 MB | NPU
111
- | FFNet-40S | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 24.885 ms | 1 - 317 MB | NPU
112
- | FFNet-40S | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 134.647 ms | 3 - 213 MB | NPU
113
- | FFNet-40S | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 36.505 ms | 2 - 5 MB | NPU
114
- | FFNet-40S | TFLITE | float | Qualcomm® SA8775P | 48.866 ms | 2 - 213 MB | NPU
115
- | FFNet-40S | TFLITE | float | Qualcomm® QCS9075 | 61.819 ms | 0 - 56 MB | NPU
116
- | FFNet-40S | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 115.735 ms | 2 - 324 MB | NPU
117
- | FFNet-40S | TFLITE | float | Qualcomm® SA7255P | 134.647 ms | 3 - 213 MB | NPU
118
- | FFNet-40S | TFLITE | float | Qualcomm® SA8295P | 64.956 ms | 0 - 219 MB | NPU
119
- | FFNet-40S | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 17.949 ms | 1 - 240 MB | NPU
120
- | FFNet-40S | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 12.609 ms | 2 - 253 MB | NPU
121
- | FFNet-40S | TFLITE | w8a8 | Snapdragon® 8 Gen 3 Mobile | 5.444 ms | 1 - 248 MB | NPU
122
- | FFNet-40S | TFLITE | w8a8 | Qualcomm® QCS6490 | 52.251 ms | 1 - 23 MB | NPU
123
- | FFNet-40S | TFLITE | w8a8 | Qualcomm® QCS8275 (Proxy) | 20.416 ms | 1 - 191 MB | NPU
124
- | FFNet-40S | TFLITE | w8a8 | Qualcomm® QCS8550 (Proxy) | 7.616 ms | 1 - 2 MB | NPU
125
- | FFNet-40S | TFLITE | w8a8 | Qualcomm® SA8775P | 8.228 ms | 1 - 192 MB | NPU
126
- | FFNet-40S | TFLITE | w8a8 | Qualcomm® QCS9075 | 9.533 ms | 0 - 22 MB | NPU
127
- | FFNet-40S | TFLITE | w8a8 | Qualcomm® QCM6690 | 101.921 ms | 1 - 232 MB | NPU
128
- | FFNet-40S | TFLITE | w8a8 | Qualcomm® QCS8450 (Proxy) | 12.264 ms | 1 - 247 MB | NPU
129
- | FFNet-40S | TFLITE | w8a8 | Qualcomm® SA7255P | 20.416 ms | 1 - 191 MB | NPU
130
- | FFNet-40S | TFLITE | w8a8 | Qualcomm® SA8295P | 11.815 ms | 1 - 195 MB | NPU
131
- | FFNet-40S | TFLITE | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 4.047 ms | 0 - 207 MB | NPU
132
- | FFNet-40S | TFLITE | w8a8 | Snapdragon® 7 Gen 4 Mobile | 11.564 ms | 0 - 213 MB | NPU
133
- | FFNet-40S | TFLITE | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 2.941 ms | 1 - 229 MB | NPU
134
 
135
  ## License
136
  * The license for the original implementation of FFNet-40S can be found
 
15
  FFNet-40S is a "fuss-free network" that segments street scene images with per-pixel classes like road, sidewalk, and pedestrian. Trained on the Cityscapes dataset.
16
 
17
  This is based on the implementation of FFNet-40S found [here](https://github.com/Qualcomm-AI-research/FFNet).
18
+ This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/tree/v0.49.1/qai_hub_models/models/ffnet_40s) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
19
 
20
  Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) to run these models on a hosted Qualcomm® device.
21
 
 
28
 
29
  | Runtime | Precision | Chipset | SDK Versions | Download |
30
  |---|---|---|---|---|
31
+ | ONNX | float | Universal | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/ffnet_40s/releases/v0.49.1/ffnet_40s-onnx-float.zip)
32
+ | ONNX | w8a8 | Universal | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/ffnet_40s/releases/v0.49.1/ffnet_40s-onnx-w8a8.zip)
33
+ | QNN_DLC | float | Universal | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/ffnet_40s/releases/v0.49.1/ffnet_40s-qnn_dlc-float.zip)
34
+ | QNN_DLC | w8a8 | Universal | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/ffnet_40s/releases/v0.49.1/ffnet_40s-qnn_dlc-w8a8.zip)
35
+ | TFLITE | float | Universal | QAIRT 2.43, TFLite 2.17.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/ffnet_40s/releases/v0.49.1/ffnet_40s-tflite-float.zip)
36
+ | TFLITE | w8a8 | Universal | QAIRT 2.43, TFLite 2.17.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/ffnet_40s/releases/v0.49.1/ffnet_40s-tflite-w8a8.zip)
37
 
38
  For more device-specific assets and performance metrics, visit **[FFNet-40S on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/ffnet_40s)**.
39
 
40
 
41
  ### Option 2: Export with Custom Configurations
42
 
43
+ Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/tree/v0.49.1/qai_hub_models/models/ffnet_40s) Python library to compile and export the model with your own:
44
  - Custom weights (e.g., fine-tuned checkpoints)
45
  - Custom input shapes
46
  - Target device and runtime configurations
47
 
48
  This option is ideal if you need to customize the model beyond the default configuration provided here.
49
 
50
+ See our repository for [FFNet-40S on GitHub](https://github.com/qualcomm/ai-hub-models/tree/v0.49.1/qai_hub_models/models/ffnet_40s) for usage instructions.
51
 
52
  ## Model Details
53
 
 
64
  ## Performance Summary
65
  | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
66
  |---|---|---|---|---|---|---
67
+ | FFNet-40S | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 12.381 ms | 29 - 250 MB | NPU
68
+ | FFNet-40S | ONNX | float | Snapdragon® X2 Elite | 13.223 ms | 22 - 22 MB | NPU
69
+ | FFNet-40S | ONNX | float | Snapdragon® X Elite | 31.612 ms | 24 - 24 MB | NPU
70
+ | FFNet-40S | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 22.515 ms | 31 - 303 MB | NPU
71
+ | FFNet-40S | ONNX | float | Qualcomm® QCS8550 (Proxy) | 32.425 ms | 24 - 26 MB | NPU
72
+ | FFNet-40S | ONNX | float | Qualcomm® QCS9075 | 47.629 ms | 24 - 27 MB | NPU
73
+ | FFNet-40S | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 15.891 ms | 6 - 206 MB | NPU
74
+ | FFNet-40S | ONNX | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 6.547 ms | 0 - 195 MB | NPU
75
+ | FFNet-40S | ONNX | w8a8 | Snapdragon® X2 Elite | 6.949 ms | 9 - 9 MB | NPU
76
+ | FFNet-40S | ONNX | w8a8 | Snapdragon® X Elite | 10.477 ms | 8 - 8 MB | NPU
77
+ | FFNet-40S | ONNX | w8a8 | Snapdragon® 8 Gen 3 Mobile | 6.553 ms | 7 - 251 MB | NPU
78
+ | FFNet-40S | ONNX | w8a8 | Qualcomm® QCS6490 | 360.417 ms | 200 - 237 MB | CPU
79
+ | FFNet-40S | ONNX | w8a8 | Qualcomm® QCS8550 (Proxy) | 9.806 ms | 0 - 213 MB | NPU
80
+ | FFNet-40S | ONNX | w8a8 | Qualcomm® QCS9075 | 13.19 ms | 6 - 9 MB | NPU
81
+ | FFNet-40S | ONNX | w8a8 | Qualcomm® QCM6690 | 365.897 ms | 211 - 219 MB | CPU
82
+ | FFNet-40S | ONNX | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 7.732 ms | 1 - 192 MB | NPU
83
+ | FFNet-40S | ONNX | w8a8 | Snapdragon® 7 Gen 4 Mobile | 360.735 ms | 132 - 142 MB | CPU
84
+ | FFNet-40S | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 12.587 ms | 12 - 250 MB | NPU
85
+ | FFNet-40S | QNN_DLC | float | Snapdragon® X2 Elite | 14.469 ms | 24 - 24 MB | NPU
86
+ | FFNet-40S | QNN_DLC | float | Snapdragon® X Elite | 37.157 ms | 24 - 24 MB | NPU
87
+ | FFNet-40S | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 24.938 ms | 21 - 301 MB | NPU
88
+ | FFNet-40S | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 134.433 ms | 24 - 219 MB | NPU
89
+ | FFNet-40S | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 36.112 ms | 24 - 26 MB | NPU
90
+ | FFNet-40S | QNN_DLC | float | Qualcomm® SA8775P | 48.774 ms | 24 - 221 MB | NPU
91
+ | FFNet-40S | QNN_DLC | float | Qualcomm® QCS9075 | 62.409 ms | 24 - 52 MB | NPU
92
+ | FFNet-40S | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 113.799 ms | 23 - 308 MB | NPU
93
+ | FFNet-40S | QNN_DLC | float | Qualcomm® SA7255P | 134.433 ms | 24 - 219 MB | NPU
94
+ | FFNet-40S | QNN_DLC | float | Qualcomm® SA8295P | 64.921 ms | 24 - 226 MB | NPU
95
+ | FFNet-40S | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 18.019 ms | 17 - 243 MB | NPU
96
+ | FFNet-40S | QNN_DLC | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 5.22 ms | 6 - 236 MB | NPU
97
+ | FFNet-40S | QNN_DLC | w8a8 | Snapdragon® X2 Elite | 6.198 ms | 6 - 6 MB | NPU
98
+ | FFNet-40S | QNN_DLC | w8a8 | Snapdragon® X Elite | 15.962 ms | 6 - 6 MB | NPU
99
+ | FFNet-40S | QNN_DLC | w8a8 | Snapdragon® 8 Gen 3 Mobile | 10.564 ms | 6 - 249 MB | NPU
100
+ | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® QCS6490 | 66.788 ms | 4 - 12 MB | NPU
101
+ | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® QCS8275 (Proxy) | 32.896 ms | 6 - 198 MB | NPU
102
+ | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® QCS8550 (Proxy) | 15.115 ms | 6 - 32 MB | NPU
103
+ | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® SA8775P | 15.763 ms | 6 - 198 MB | NPU
104
+ | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® QCS9075 | 18.677 ms | 6 - 14 MB | NPU
105
+ | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® QCM6690 | 126.535 ms | 6 - 238 MB | NPU
106
+ | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® QCS8450 (Proxy) | 21.522 ms | 6 - 249 MB | NPU
107
+ | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® SA7255P | 32.896 ms | 6 - 198 MB | NPU
108
+ | FFNet-40S | QNN_DLC | w8a8 | Qualcomm® SA8295P | 20.181 ms | 6 - 201 MB | NPU
109
+ | FFNet-40S | QNN_DLC | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 7.225 ms | 6 - 214 MB | NPU
110
+ | FFNet-40S | QNN_DLC | w8a8 | Snapdragon® 7 Gen 4 Mobile | 19.056 ms | 6 - 221 MB | NPU
111
+ | FFNet-40S | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 12.641 ms | 2 - 252 MB | NPU
112
+ | FFNet-40S | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 24.815 ms | 2 - 318 MB | NPU
113
+ | FFNet-40S | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 134.663 ms | 3 - 213 MB | NPU
114
+ | FFNet-40S | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 36.159 ms | 2 - 5 MB | NPU
115
+ | FFNet-40S | TFLITE | float | Qualcomm® SA8775P | 48.855 ms | 0 - 210 MB | NPU
116
+ | FFNet-40S | TFLITE | float | Qualcomm® QCS9075 | 61.979 ms | 0 - 56 MB | NPU
117
+ | FFNet-40S | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 117.534 ms | 2 - 323 MB | NPU
118
+ | FFNet-40S | TFLITE | float | Qualcomm® SA7255P | 134.663 ms | 3 - 213 MB | NPU
119
+ | FFNet-40S | TFLITE | float | Qualcomm® SA8295P | 64.926 ms | 2 - 220 MB | NPU
120
+ | FFNet-40S | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 17.991 ms | 1 - 242 MB | NPU
121
+ | FFNet-40S | TFLITE | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 2.94 ms | 1 - 230 MB | NPU
122
+ | FFNet-40S | TFLITE | w8a8 | Snapdragon® 8 Gen 3 Mobile | 5.438 ms | 0 - 245 MB | NPU
123
+ | FFNet-40S | TFLITE | w8a8 | Qualcomm® QCS6490 | 52.249 ms | 0 - 23 MB | NPU
124
+ | FFNet-40S | TFLITE | w8a8 | Qualcomm® QCS8275 (Proxy) | 20.435 ms | 1 - 191 MB | NPU
125
+ | FFNet-40S | TFLITE | w8a8 | Qualcomm® QCS8550 (Proxy) | 7.611 ms | 1 - 3 MB | NPU
126
+ | FFNet-40S | TFLITE | w8a8 | Qualcomm® SA8775P | 8.234 ms | 1 - 192 MB | NPU
127
+ | FFNet-40S | TFLITE | w8a8 | Qualcomm® QCS9075 | 9.565 ms | 0 - 22 MB | NPU
128
+ | FFNet-40S | TFLITE | w8a8 | Qualcomm® QCM6690 | 102.559 ms | 1 - 232 MB | NPU
129
+ | FFNet-40S | TFLITE | w8a8 | Qualcomm® QCS8450 (Proxy) | 12.176 ms | 0 - 246 MB | NPU
130
+ | FFNet-40S | TFLITE | w8a8 | Qualcomm® SA7255P | 20.435 ms | 1 - 191 MB | NPU
131
+ | FFNet-40S | TFLITE | w8a8 | Qualcomm® SA8295P | 11.778 ms | 1 - 196 MB | NPU
132
+ | FFNet-40S | TFLITE | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 4.071 ms | 0 - 207 MB | NPU
133
+ | FFNet-40S | TFLITE | w8a8 | Snapdragon® 7 Gen 4 Mobile | 11.585 ms | 0 - 212 MB | NPU
134
 
135
  ## License
136
  * The license for the original implementation of FFNet-40S can be found