v0.34.0
Browse filesSee https://github.com/quic/ai-hub-models/releases/v0.34.0 for changelog.
README.md
CHANGED
|
@@ -19,7 +19,11 @@ YoloV7 is a machine learning model that predicts bounding boxes and classes of o
|
|
| 19 |
This model is an implementation of Yolo-v7 found [here](https://github.com/WongKinYiu/yolov7/).
|
| 20 |
|
| 21 |
|
| 22 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 23 |
|
| 24 |
### Model Details
|
| 25 |
|
|
@@ -34,87 +38,255 @@ This model is an implementation of Yolo-v7 found [here](https://github.com/WongK
|
|
| 34 |
|
| 35 |
| Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
|
| 36 |
|---|---|---|---|---|---|---|---|---|
|
| 37 |
-
| Yolo-v7 | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE |
|
| 38 |
| Yolo-v7 | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 22.796 ms | 0 - 119 MB | NPU | -- |
|
| 39 |
-
| Yolo-v7 | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 20.598 ms | 1 -
|
| 40 |
| Yolo-v7 | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 13.777 ms | 5 - 44 MB | NPU | -- |
|
| 41 |
-
| Yolo-v7 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 17.
|
| 42 |
| Yolo-v7 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 8.761 ms | 5 - 22 MB | NPU | -- |
|
| 43 |
-
| Yolo-v7 | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 18.
|
| 44 |
| Yolo-v7 | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 10.511 ms | 1 - 123 MB | NPU | -- |
|
| 45 |
-
| Yolo-v7 | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE |
|
| 46 |
| Yolo-v7 | float | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 22.796 ms | 0 - 119 MB | NPU | -- |
|
| 47 |
-
| Yolo-v7 | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 17.
|
| 48 |
| Yolo-v7 | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 8.792 ms | 0 - 97 MB | NPU | -- |
|
| 49 |
-
| Yolo-v7 | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 22.
|
| 50 |
| Yolo-v7 | float | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 13.626 ms | 2 - 32 MB | NPU | -- |
|
| 51 |
-
| Yolo-v7 | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 17.
|
| 52 |
| Yolo-v7 | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 8.771 ms | 0 - 78 MB | NPU | -- |
|
| 53 |
-
| Yolo-v7 | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 18.
|
| 54 |
| Yolo-v7 | float | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 10.511 ms | 1 - 123 MB | NPU | -- |
|
| 55 |
-
| Yolo-v7 | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 17.
|
| 56 |
| Yolo-v7 | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN_DLC | 8.78 ms | 0 - 88 MB | NPU | -- |
|
| 57 |
-
| Yolo-v7 | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 11.
|
| 58 |
-
| Yolo-v7 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 12.
|
| 59 |
| Yolo-v7 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 5.823 ms | 5 - 275 MB | NPU | -- |
|
| 60 |
-
| Yolo-v7 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 7.
|
| 61 |
-
| Yolo-v7 | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE |
|
| 62 |
| Yolo-v7 | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN_DLC | 5.908 ms | 5 - 136 MB | NPU | -- |
|
| 63 |
-
| Yolo-v7 | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX |
|
| 64 |
| Yolo-v7 | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 9.64 ms | 87 - 87 MB | NPU | -- |
|
| 65 |
-
| Yolo-v7 | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 13.
|
| 66 |
-
| Yolo-v7 | w8a16 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC |
|
| 67 |
-
| Yolo-v7 | w8a16 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 10.
|
| 68 |
-
| Yolo-v7 | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 9.
|
| 69 |
-
| Yolo-v7 | w8a16 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC |
|
| 70 |
-
| Yolo-v7 | w8a16 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN_DLC |
|
| 71 |
-
| Yolo-v7 | w8a16 | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC |
|
| 72 |
-
| Yolo-v7 | w8a16 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 9.
|
| 73 |
-
| Yolo-v7 | w8a16 | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 11.
|
| 74 |
-
| Yolo-v7 | w8a16 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 9.
|
| 75 |
-
| Yolo-v7 | w8a16 | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC |
|
| 76 |
-
| Yolo-v7 | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN_DLC | 9.
|
| 77 |
-
| Yolo-v7 | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 6.
|
| 78 |
-
| Yolo-v7 | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 6.
|
| 79 |
-
| Yolo-v7 | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 4.
|
| 80 |
-
| Yolo-v7 | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN_DLC | 6.
|
| 81 |
-
| Yolo-v7 | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 4.
|
| 82 |
-
| Yolo-v7 | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 10.
|
| 83 |
-
| Yolo-v7 | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 7.
|
| 84 |
-
| Yolo-v7 | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 4.
|
| 85 |
-
| Yolo-v7 | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 4.
|
| 86 |
-
| Yolo-v7 | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 2.
|
| 87 |
-
| Yolo-v7 | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 2.
|
| 88 |
-
| Yolo-v7 | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 2.
|
| 89 |
-
| Yolo-v7 | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 2.
|
| 90 |
-
| Yolo-v7 | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 2.
|
| 91 |
-
| Yolo-v7 | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 2.
|
| 92 |
-
| Yolo-v7 | w8a8 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | TFLITE | 9.
|
| 93 |
-
| Yolo-v7 | w8a8 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN_DLC | 7.
|
| 94 |
-
| Yolo-v7 | w8a8 | RB5 (Proxy) | Qualcomm® QCS8250 (Proxy) | TFLITE |
|
| 95 |
-
| Yolo-v7 | w8a8 | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 4.
|
| 96 |
-
| Yolo-v7 | w8a8 | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 4.
|
| 97 |
-
| Yolo-v7 | w8a8 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 2.126 ms | 0 -
|
| 98 |
-
| Yolo-v7 | w8a8 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 2.
|
| 99 |
-
| Yolo-v7 | w8a8 | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 3.
|
| 100 |
-
| Yolo-v7 | w8a8 | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 3.
|
| 101 |
-
| Yolo-v7 | w8a8 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 2.
|
| 102 |
-
| Yolo-v7 | w8a8 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 2.081 ms |
|
| 103 |
-
| Yolo-v7 | w8a8 | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 2.
|
| 104 |
-
| Yolo-v7 | w8a8 | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 2.
|
| 105 |
-
| Yolo-v7 | w8a8 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 2.
|
| 106 |
-
| Yolo-v7 | w8a8 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN_DLC | 2.
|
| 107 |
-
| Yolo-v7 | w8a8 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 3.
|
| 108 |
-
| Yolo-v7 | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 1.
|
| 109 |
-
| Yolo-v7 | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 1.
|
| 110 |
-
| Yolo-v7 | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX |
|
| 111 |
-
| Yolo-v7 | w8a8 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 1.
|
| 112 |
-
| Yolo-v7 | w8a8 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN_DLC | 1.
|
| 113 |
-
| Yolo-v7 | w8a8 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 2.
|
| 114 |
-
| Yolo-v7 | w8a8 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 2.521 ms |
|
| 115 |
-
| Yolo-v7 | w8a8 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 4.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 116 |
|
|
|
|
|
|
|
| 117 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 118 |
|
| 119 |
|
| 120 |
## License
|
|
@@ -131,26 +303,7 @@ This model is an implementation of Yolo-v7 found [here](https://github.com/WongK
|
|
| 131 |
|
| 132 |
|
| 133 |
## Community
|
| 134 |
-
* Join [our AI Hub Slack community](https://qualcomm
|
| 135 |
* For questions or feedback please [reach out to us](mailto:[email protected]).
|
| 136 |
|
| 137 |
-
|
| 138 |
-
|
| 139 |
-
Model may not be used for or in connection with any of the following applications:
|
| 140 |
-
|
| 141 |
-
- Accessing essential private and public services and benefits;
|
| 142 |
-
- Administration of justice and democratic processes;
|
| 143 |
-
- Assessing or recognizing the emotional state of a person;
|
| 144 |
-
- Biometric and biometrics-based systems, including categorization of persons based on sensitive characteristics;
|
| 145 |
-
- Education and vocational training;
|
| 146 |
-
- Employment and workers management;
|
| 147 |
-
- Exploitation of the vulnerabilities of persons resulting in harmful behavior;
|
| 148 |
-
- General purpose social scoring;
|
| 149 |
-
- Law enforcement;
|
| 150 |
-
- Management and operation of critical infrastructure;
|
| 151 |
-
- Migration, asylum and border control management;
|
| 152 |
-
- Predictive policing;
|
| 153 |
-
- Real-time remote biometric identification in public spaces;
|
| 154 |
-
- Recommender systems of social media platforms;
|
| 155 |
-
- Scraping of facial images (from the internet or otherwise); and/or
|
| 156 |
-
- Subliminal manipulation
|
|
|
|
| 19 |
This model is an implementation of Yolo-v7 found [here](https://github.com/WongKinYiu/yolov7/).
|
| 20 |
|
| 21 |
|
| 22 |
+
This repository provides scripts to run Yolo-v7 on Qualcomm® devices.
|
| 23 |
+
More details on model performance across various devices, can be found
|
| 24 |
+
[here](https://aihub.qualcomm.com/models/yolov7).
|
| 25 |
+
|
| 26 |
+
**WARNING**: The model assets are not readily available for download due to licensing restrictions.
|
| 27 |
|
| 28 |
### Model Details
|
| 29 |
|
|
|
|
| 38 |
|
| 39 |
| Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
|
| 40 |
|---|---|---|---|---|---|---|---|---|
|
| 41 |
+
| Yolo-v7 | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 35.08 ms | 1 - 27 MB | NPU | -- |
|
| 42 |
| Yolo-v7 | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 22.796 ms | 0 - 119 MB | NPU | -- |
|
| 43 |
+
| Yolo-v7 | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 20.598 ms | 1 - 43 MB | NPU | -- |
|
| 44 |
| Yolo-v7 | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 13.777 ms | 5 - 44 MB | NPU | -- |
|
| 45 |
+
| Yolo-v7 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 17.457 ms | 1 - 12 MB | NPU | -- |
|
| 46 |
| Yolo-v7 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 8.761 ms | 5 - 22 MB | NPU | -- |
|
| 47 |
+
| Yolo-v7 | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 18.727 ms | 1 - 28 MB | NPU | -- |
|
| 48 |
| Yolo-v7 | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 10.511 ms | 1 - 123 MB | NPU | -- |
|
| 49 |
+
| Yolo-v7 | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 35.08 ms | 1 - 27 MB | NPU | -- |
|
| 50 |
| Yolo-v7 | float | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 22.796 ms | 0 - 119 MB | NPU | -- |
|
| 51 |
+
| Yolo-v7 | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 17.67 ms | 1 - 17 MB | NPU | -- |
|
| 52 |
| Yolo-v7 | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 8.792 ms | 0 - 97 MB | NPU | -- |
|
| 53 |
+
| Yolo-v7 | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 22.117 ms | 1 - 32 MB | NPU | -- |
|
| 54 |
| Yolo-v7 | float | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 13.626 ms | 2 - 32 MB | NPU | -- |
|
| 55 |
+
| Yolo-v7 | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 17.735 ms | 0 - 13 MB | NPU | -- |
|
| 56 |
| Yolo-v7 | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 8.771 ms | 0 - 78 MB | NPU | -- |
|
| 57 |
+
| Yolo-v7 | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 18.727 ms | 1 - 28 MB | NPU | -- |
|
| 58 |
| Yolo-v7 | float | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 10.511 ms | 1 - 123 MB | NPU | -- |
|
| 59 |
+
| Yolo-v7 | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 17.695 ms | 1 - 12 MB | NPU | -- |
|
| 60 |
| Yolo-v7 | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN_DLC | 8.78 ms | 0 - 88 MB | NPU | -- |
|
| 61 |
+
| Yolo-v7 | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 11.093 ms | 0 - 25 MB | NPU | -- |
|
| 62 |
+
| Yolo-v7 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 12.915 ms | 1 - 41 MB | NPU | -- |
|
| 63 |
| Yolo-v7 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 5.823 ms | 5 - 275 MB | NPU | -- |
|
| 64 |
+
| Yolo-v7 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 7.854 ms | 5 - 214 MB | NPU | -- |
|
| 65 |
+
| Yolo-v7 | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 10.61 ms | 0 - 33 MB | NPU | -- |
|
| 66 |
| Yolo-v7 | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN_DLC | 5.908 ms | 5 - 136 MB | NPU | -- |
|
| 67 |
+
| Yolo-v7 | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 8.216 ms | 5 - 128 MB | NPU | -- |
|
| 68 |
| Yolo-v7 | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 9.64 ms | 87 - 87 MB | NPU | -- |
|
| 69 |
+
| Yolo-v7 | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 13.065 ms | 9 - 9 MB | NPU | -- |
|
| 70 |
+
| Yolo-v7 | w8a16 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 18.225 ms | 0 - 34 MB | NPU | -- |
|
| 71 |
+
| Yolo-v7 | w8a16 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 10.796 ms | 2 - 54 MB | NPU | -- |
|
| 72 |
+
| Yolo-v7 | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 9.348 ms | 2 - 13 MB | NPU | -- |
|
| 73 |
+
| Yolo-v7 | w8a16 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 9.842 ms | 0 - 35 MB | NPU | -- |
|
| 74 |
+
| Yolo-v7 | w8a16 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN_DLC | 29.634 ms | 0 - 33 MB | NPU | -- |
|
| 75 |
+
| Yolo-v7 | w8a16 | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 18.225 ms | 0 - 34 MB | NPU | -- |
|
| 76 |
+
| Yolo-v7 | w8a16 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 9.303 ms | 2 - 15 MB | NPU | -- |
|
| 77 |
+
| Yolo-v7 | w8a16 | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 11.662 ms | 1 - 43 MB | NPU | -- |
|
| 78 |
+
| Yolo-v7 | w8a16 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 9.395 ms | 2 - 13 MB | NPU | -- |
|
| 79 |
+
| Yolo-v7 | w8a16 | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 9.842 ms | 0 - 35 MB | NPU | -- |
|
| 80 |
+
| Yolo-v7 | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN_DLC | 9.384 ms | 2 - 14 MB | NPU | -- |
|
| 81 |
+
| Yolo-v7 | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 6.248 ms | 0 - 45 MB | NPU | -- |
|
| 82 |
+
| Yolo-v7 | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 6.215 ms | 2 - 53 MB | NPU | -- |
|
| 83 |
+
| Yolo-v7 | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 4.175 ms | 2 - 224 MB | NPU | -- |
|
| 84 |
+
| Yolo-v7 | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN_DLC | 6.729 ms | 2 - 55 MB | NPU | -- |
|
| 85 |
+
| Yolo-v7 | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 4.099 ms | 1 - 154 MB | NPU | -- |
|
| 86 |
+
| Yolo-v7 | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 10.434 ms | 0 - 0 MB | NPU | -- |
|
| 87 |
+
| Yolo-v7 | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 7.108 ms | 5 - 5 MB | NPU | -- |
|
| 88 |
+
| Yolo-v7 | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 4.706 ms | 0 - 23 MB | NPU | -- |
|
| 89 |
+
| Yolo-v7 | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 4.626 ms | 1 - 28 MB | NPU | -- |
|
| 90 |
+
| Yolo-v7 | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 2.631 ms | 0 - 37 MB | NPU | -- |
|
| 91 |
+
| Yolo-v7 | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 2.939 ms | 1 - 37 MB | NPU | -- |
|
| 92 |
+
| Yolo-v7 | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 2.105 ms | 0 - 10 MB | NPU | -- |
|
| 93 |
+
| Yolo-v7 | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 2.073 ms | 1 - 11 MB | NPU | -- |
|
| 94 |
+
| Yolo-v7 | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 2.539 ms | 0 - 25 MB | NPU | -- |
|
| 95 |
+
| Yolo-v7 | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 2.471 ms | 1 - 28 MB | NPU | -- |
|
| 96 |
+
| Yolo-v7 | w8a8 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | TFLITE | 9.931 ms | 0 - 36 MB | NPU | -- |
|
| 97 |
+
| Yolo-v7 | w8a8 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN_DLC | 7.715 ms | 0 - 33 MB | NPU | -- |
|
| 98 |
+
| Yolo-v7 | w8a8 | RB5 (Proxy) | Qualcomm® QCS8250 (Proxy) | TFLITE | 57.957 ms | 15 - 35 MB | GPU | -- |
|
| 99 |
+
| Yolo-v7 | w8a8 | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 4.706 ms | 0 - 23 MB | NPU | -- |
|
| 100 |
+
| Yolo-v7 | w8a8 | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 4.626 ms | 1 - 28 MB | NPU | -- |
|
| 101 |
+
| Yolo-v7 | w8a8 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 2.126 ms | 0 - 29 MB | NPU | -- |
|
| 102 |
+
| Yolo-v7 | w8a8 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 2.071 ms | 1 - 14 MB | NPU | -- |
|
| 103 |
+
| Yolo-v7 | w8a8 | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 3.437 ms | 0 - 25 MB | NPU | -- |
|
| 104 |
+
| Yolo-v7 | w8a8 | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 3.447 ms | 1 - 32 MB | NPU | -- |
|
| 105 |
+
| Yolo-v7 | w8a8 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 2.124 ms | 0 - 34 MB | NPU | -- |
|
| 106 |
+
| Yolo-v7 | w8a8 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 2.081 ms | 1 - 13 MB | NPU | -- |
|
| 107 |
+
| Yolo-v7 | w8a8 | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 2.539 ms | 0 - 25 MB | NPU | -- |
|
| 108 |
+
| Yolo-v7 | w8a8 | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 2.471 ms | 1 - 28 MB | NPU | -- |
|
| 109 |
+
| Yolo-v7 | w8a8 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 2.123 ms | 0 - 40 MB | NPU | -- |
|
| 110 |
+
| Yolo-v7 | w8a8 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN_DLC | 2.08 ms | 1 - 13 MB | NPU | -- |
|
| 111 |
+
| Yolo-v7 | w8a8 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 3.932 ms | 0 - 59 MB | NPU | -- |
|
| 112 |
+
| Yolo-v7 | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 1.419 ms | 0 - 43 MB | NPU | -- |
|
| 113 |
+
| Yolo-v7 | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 1.423 ms | 19 - 62 MB | NPU | -- |
|
| 114 |
+
| Yolo-v7 | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 2.783 ms | 0 - 254 MB | NPU | -- |
|
| 115 |
+
| Yolo-v7 | w8a8 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 1.36 ms | 0 - 30 MB | NPU | -- |
|
| 116 |
+
| Yolo-v7 | w8a8 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN_DLC | 1.344 ms | 1 - 34 MB | NPU | -- |
|
| 117 |
+
| Yolo-v7 | w8a8 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 2.225 ms | 1 - 152 MB | NPU | -- |
|
| 118 |
+
| Yolo-v7 | w8a8 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 2.521 ms | 20 - 20 MB | NPU | -- |
|
| 119 |
+
| Yolo-v7 | w8a8 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 4.282 ms | 5 - 5 MB | NPU | -- |
|
| 120 |
+
|
| 121 |
+
|
| 122 |
+
|
| 123 |
+
|
| 124 |
+
## Installation
|
| 125 |
+
|
| 126 |
+
|
| 127 |
+
Install the package via pip:
|
| 128 |
+
```bash
|
| 129 |
+
pip install "qai-hub-models[yolov7]"
|
| 130 |
+
```
|
| 131 |
+
|
| 132 |
+
|
| 133 |
+
## Configure Qualcomm® AI Hub to run this model on a cloud-hosted device
|
| 134 |
+
|
| 135 |
+
Sign-in to [Qualcomm® AI Hub](https://app.aihub.qualcomm.com/) with your
|
| 136 |
+
Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.
|
| 137 |
+
|
| 138 |
+
With this API token, you can configure your client to run models on the cloud
|
| 139 |
+
hosted devices.
|
| 140 |
+
```bash
|
| 141 |
+
qai-hub configure --api_token API_TOKEN
|
| 142 |
+
```
|
| 143 |
+
Navigate to [docs](https://app.aihub.qualcomm.com/docs/) for more information.
|
| 144 |
+
|
| 145 |
+
|
| 146 |
+
|
| 147 |
+
## Demo off target
|
| 148 |
+
|
| 149 |
+
The package contains a simple end-to-end demo that downloads pre-trained
|
| 150 |
+
weights and runs this model on a sample input.
|
| 151 |
+
|
| 152 |
+
```bash
|
| 153 |
+
python -m qai_hub_models.models.yolov7.demo
|
| 154 |
+
```
|
| 155 |
+
|
| 156 |
+
The above demo runs a reference implementation of pre-processing, model
|
| 157 |
+
inference, and post processing.
|
| 158 |
+
|
| 159 |
+
**NOTE**: If you want running in a Jupyter Notebook or Google Colab like
|
| 160 |
+
environment, please add the following to your cell (instead of the above).
|
| 161 |
+
```
|
| 162 |
+
%run -m qai_hub_models.models.yolov7.demo
|
| 163 |
+
```
|
| 164 |
+
|
| 165 |
+
|
| 166 |
+
### Run model on a cloud-hosted device
|
| 167 |
+
|
| 168 |
+
In addition to the demo, you can also run the model on a cloud-hosted Qualcomm®
|
| 169 |
+
device. This script does the following:
|
| 170 |
+
* Performance check on-device on a cloud-hosted device
|
| 171 |
+
* Downloads compiled assets that can be deployed on-device for Android.
|
| 172 |
+
* Accuracy check between PyTorch and on-device outputs.
|
| 173 |
+
|
| 174 |
+
```bash
|
| 175 |
+
python -m qai_hub_models.models.yolov7.export
|
| 176 |
+
```
|
| 177 |
+
|
| 178 |
+
|
| 179 |
+
|
| 180 |
+
## How does this work?
|
| 181 |
+
|
| 182 |
+
This [export script](https://aihub.qualcomm.com/models/yolov7/qai_hub_models/models/Yolo-v7/export.py)
|
| 183 |
+
leverages [Qualcomm® AI Hub](https://aihub.qualcomm.com/) to optimize, validate, and deploy this model
|
| 184 |
+
on-device. Lets go through each step below in detail:
|
| 185 |
+
|
| 186 |
+
Step 1: **Compile model for on-device deployment**
|
| 187 |
+
|
| 188 |
+
To compile a PyTorch model for on-device deployment, we first trace the model
|
| 189 |
+
in memory using the `jit.trace` and then call the `submit_compile_job` API.
|
| 190 |
+
|
| 191 |
+
```python
|
| 192 |
+
import torch
|
| 193 |
|
| 194 |
+
import qai_hub as hub
|
| 195 |
+
from qai_hub_models.models.yolov7 import Model
|
| 196 |
|
| 197 |
+
# Load the model
|
| 198 |
+
torch_model = Model.from_pretrained()
|
| 199 |
+
|
| 200 |
+
# Device
|
| 201 |
+
device = hub.Device("Samsung Galaxy S24")
|
| 202 |
+
|
| 203 |
+
# Trace model
|
| 204 |
+
input_shape = torch_model.get_input_spec()
|
| 205 |
+
sample_inputs = torch_model.sample_inputs()
|
| 206 |
+
|
| 207 |
+
pt_model = torch.jit.trace(torch_model, [torch.tensor(data[0]) for _, data in sample_inputs.items()])
|
| 208 |
+
|
| 209 |
+
# Compile model on a specific device
|
| 210 |
+
compile_job = hub.submit_compile_job(
|
| 211 |
+
model=pt_model,
|
| 212 |
+
device=device,
|
| 213 |
+
input_specs=torch_model.get_input_spec(),
|
| 214 |
+
)
|
| 215 |
+
|
| 216 |
+
# Get target model to run on-device
|
| 217 |
+
target_model = compile_job.get_target_model()
|
| 218 |
+
|
| 219 |
+
```
|
| 220 |
+
|
| 221 |
+
|
| 222 |
+
Step 2: **Performance profiling on cloud-hosted device**
|
| 223 |
+
|
| 224 |
+
After compiling models from step 1. Models can be profiled model on-device using the
|
| 225 |
+
`target_model`. Note that this scripts runs the model on a device automatically
|
| 226 |
+
provisioned in the cloud. Once the job is submitted, you can navigate to a
|
| 227 |
+
provided job URL to view a variety of on-device performance metrics.
|
| 228 |
+
```python
|
| 229 |
+
profile_job = hub.submit_profile_job(
|
| 230 |
+
model=target_model,
|
| 231 |
+
device=device,
|
| 232 |
+
)
|
| 233 |
+
|
| 234 |
+
```
|
| 235 |
+
|
| 236 |
+
Step 3: **Verify on-device accuracy**
|
| 237 |
+
|
| 238 |
+
To verify the accuracy of the model on-device, you can run on-device inference
|
| 239 |
+
on sample input data on the same cloud hosted device.
|
| 240 |
+
```python
|
| 241 |
+
input_data = torch_model.sample_inputs()
|
| 242 |
+
inference_job = hub.submit_inference_job(
|
| 243 |
+
model=target_model,
|
| 244 |
+
device=device,
|
| 245 |
+
inputs=input_data,
|
| 246 |
+
)
|
| 247 |
+
on_device_output = inference_job.download_output_data()
|
| 248 |
+
|
| 249 |
+
```
|
| 250 |
+
With the output of the model, you can compute like PSNR, relative errors or
|
| 251 |
+
spot check the output with expected output.
|
| 252 |
+
|
| 253 |
+
**Note**: This on-device profiling and inference requires access to Qualcomm®
|
| 254 |
+
AI Hub. [Sign up for access](https://myaccount.qualcomm.com/signup).
|
| 255 |
+
|
| 256 |
+
|
| 257 |
+
|
| 258 |
+
## Run demo on a cloud-hosted device
|
| 259 |
+
|
| 260 |
+
You can also run the demo on-device.
|
| 261 |
+
|
| 262 |
+
```bash
|
| 263 |
+
python -m qai_hub_models.models.yolov7.demo --eval-mode on-device
|
| 264 |
+
```
|
| 265 |
+
|
| 266 |
+
**NOTE**: If you want running in a Jupyter Notebook or Google Colab like
|
| 267 |
+
environment, please add the following to your cell (instead of the above).
|
| 268 |
+
```
|
| 269 |
+
%run -m qai_hub_models.models.yolov7.demo -- --eval-mode on-device
|
| 270 |
+
```
|
| 271 |
+
|
| 272 |
+
|
| 273 |
+
## Deploying compiled model to Android
|
| 274 |
+
|
| 275 |
+
|
| 276 |
+
The models can be deployed using multiple runtimes:
|
| 277 |
+
- TensorFlow Lite (`.tflite` export): [This
|
| 278 |
+
tutorial](https://www.tensorflow.org/lite/android/quickstart) provides a
|
| 279 |
+
guide to deploy the .tflite model in an Android application.
|
| 280 |
+
|
| 281 |
+
|
| 282 |
+
- QNN (`.so` export ): This [sample
|
| 283 |
+
app](https://docs.qualcomm.com/bundle/publicresource/topics/80-63442-50/sample_app.html)
|
| 284 |
+
provides instructions on how to use the `.so` shared library in an Android application.
|
| 285 |
+
|
| 286 |
+
|
| 287 |
+
## View on Qualcomm® AI Hub
|
| 288 |
+
Get more details on Yolo-v7's performance across various devices [here](https://aihub.qualcomm.com/models/yolov7).
|
| 289 |
+
Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)
|
| 290 |
|
| 291 |
|
| 292 |
## License
|
|
|
|
| 303 |
|
| 304 |
|
| 305 |
## Community
|
| 306 |
+
* Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI.
|
| 307 |
* For questions or feedback please [reach out to us](mailto:[email protected]).
|
| 308 |
|
| 309 |
+
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|