Yolo-v7: Optimized for Mobile Deployment
Real-time object detection optimized for mobile and edge
YoloV7 is a machine learning model that predicts bounding boxes and classes of objects in an image.
This model is an implementation of Yolo-v7 found here.
More details on model performance across various devices, can be found here.
Model Details
- Model Type: Model_use_case.object_detection
- Model Stats:
- Model checkpoint: YoloV7 Tiny
- Input resolution: 640x640
- Number of parameters: 6.39M
- Model size (float): 24.4 MB
- Model size (w8a8): 6.23 MB
| Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model |
|---|---|---|---|---|---|---|---|---|
| Yolo-v7 | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 34.894 ms | 1 - 24 MB | NPU | -- |
| Yolo-v7 | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN | 23.761 ms | 3 - 13 MB | NPU | -- |
| Yolo-v7 | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 20.994 ms | 1 - 41 MB | NPU | -- |
| Yolo-v7 | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN | 12.754 ms | 5 - 41 MB | NPU | -- |
| Yolo-v7 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 17.584 ms | 1 - 11 MB | NPU | -- |
| Yolo-v7 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 9.153 ms | 5 - 7 MB | NPU | -- |
| Yolo-v7 | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 18.779 ms | 1 - 25 MB | NPU | -- |
| Yolo-v7 | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN | 11.054 ms | 2 - 16 MB | NPU | -- |
| Yolo-v7 | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 34.894 ms | 1 - 24 MB | NPU | -- |
| Yolo-v7 | float | SA7255P ADP | Qualcomm® SA7255P | QNN | 23.761 ms | 3 - 13 MB | NPU | -- |
| Yolo-v7 | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 17.464 ms | 2 - 16 MB | NPU | -- |
| Yolo-v7 | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN | 9.31 ms | 5 - 7 MB | NPU | -- |
| Yolo-v7 | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 22.321 ms | 0 - 29 MB | NPU | -- |
| Yolo-v7 | float | SA8295P ADP | Qualcomm® SA8295P | QNN | 12.292 ms | 0 - 18 MB | NPU | -- |
| Yolo-v7 | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 17.551 ms | 1 - 11 MB | NPU | -- |
| Yolo-v7 | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN | 9.003 ms | 5 - 8 MB | NPU | -- |
| Yolo-v7 | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 18.779 ms | 1 - 25 MB | NPU | -- |
| Yolo-v7 | float | SA8775P ADP | Qualcomm® SA8775P | QNN | 11.054 ms | 2 - 16 MB | NPU | -- |
| Yolo-v7 | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 17.611 ms | 1 - 12 MB | NPU | -- |
| Yolo-v7 | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 9.018 ms | 5 - 20 MB | NPU | -- |
| Yolo-v7 | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 10.507 ms | 2 - 47 MB | NPU | -- |
| Yolo-v7 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 12.847 ms | 1 - 38 MB | NPU | -- |
| Yolo-v7 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 6.465 ms | 5 - 60 MB | NPU | -- |
| Yolo-v7 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 8.398 ms | 7 - 68 MB | NPU | -- |
| Yolo-v7 | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 12.37 ms | 0 - 30 MB | NPU | -- |
| Yolo-v7 | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN | 5.154 ms | 5 - 55 MB | NPU | -- |
| Yolo-v7 | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 9.706 ms | 5 - 5 MB | NPU | -- |
| Yolo-v7 | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 11.47 ms | 8 - 8 MB | NPU | -- |
| Yolo-v7 | w8a16 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN | 23.364 ms | 0 - 10 MB | NPU | -- |
| Yolo-v7 | w8a16 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN | 13.748 ms | 2 - 53 MB | NPU | -- |
| Yolo-v7 | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 12.642 ms | 2 - 4 MB | NPU | -- |
| Yolo-v7 | w8a16 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN | 13.269 ms | 2 - 16 MB | NPU | -- |
| Yolo-v7 | w8a16 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN | 29.628 ms | 2 - 13 MB | NPU | -- |
| Yolo-v7 | w8a16 | SA7255P ADP | Qualcomm® SA7255P | QNN | 23.364 ms | 0 - 10 MB | NPU | -- |
| Yolo-v7 | w8a16 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN | 12.614 ms | 2 - 5 MB | NPU | -- |
| Yolo-v7 | w8a16 | SA8295P ADP | Qualcomm® SA8295P | QNN | 11.806 ms | 2 - 7 MB | NPU | -- |
| Yolo-v7 | w8a16 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN | 12.693 ms | 2 - 5 MB | NPU | -- |
| Yolo-v7 | w8a16 | SA8775P ADP | Qualcomm® SA8775P | QNN | 13.269 ms | 2 - 16 MB | NPU | -- |
| Yolo-v7 | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 12.678 ms | 2 - 15 MB | NPU | -- |
| Yolo-v7 | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 6.342 ms | 0 - 31 MB | NPU | -- |
| Yolo-v7 | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 8.518 ms | 2 - 50 MB | NPU | -- |
| Yolo-v7 | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 4.554 ms | 2 - 69 MB | NPU | -- |
| Yolo-v7 | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN | 6.695 ms | 2 - 43 MB | NPU | -- |
| Yolo-v7 | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 4.397 ms | 5 - 67 MB | NPU | -- |
| Yolo-v7 | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 13.369 ms | 2 - 2 MB | NPU | -- |
| Yolo-v7 | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 7.453 ms | 5 - 5 MB | NPU | -- |
| Yolo-v7 | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 5.323 ms | 0 - 21 MB | NPU | -- |
| Yolo-v7 | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN | 5.536 ms | 1 - 11 MB | NPU | -- |
| Yolo-v7 | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 3.128 ms | 0 - 37 MB | NPU | -- |
| Yolo-v7 | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN | 4.232 ms | 1 - 39 MB | NPU | -- |
| Yolo-v7 | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 2.526 ms | 0 - 29 MB | NPU | -- |
| Yolo-v7 | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 2.753 ms | 1 - 4 MB | NPU | -- |
| Yolo-v7 | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 2.924 ms | 0 - 24 MB | NPU | -- |
| Yolo-v7 | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN | 3.242 ms | 1 - 16 MB | NPU | -- |
| Yolo-v7 | w8a8 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | TFLITE | 9.718 ms | 0 - 34 MB | NPU | -- |
| Yolo-v7 | w8a8 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN | 7.778 ms | 1 - 16 MB | NPU | -- |
| Yolo-v7 | w8a8 | RB5 (Proxy) | Qualcomm® QCS8250 (Proxy) | TFLITE | 64.433 ms | 15 - 53 MB | GPU | -- |
| Yolo-v7 | w8a8 | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 5.323 ms | 0 - 21 MB | NPU | -- |
| Yolo-v7 | w8a8 | SA7255P ADP | Qualcomm® SA7255P | QNN | 5.536 ms | 1 - 11 MB | NPU | -- |
| Yolo-v7 | w8a8 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 2.514 ms | 0 - 30 MB | NPU | -- |
| Yolo-v7 | w8a8 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN | 2.753 ms | 1 - 3 MB | NPU | -- |
| Yolo-v7 | w8a8 | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 3.808 ms | 0 - 26 MB | NPU | -- |
| Yolo-v7 | w8a8 | SA8295P ADP | Qualcomm® SA8295P | QNN | 4.311 ms | 1 - 19 MB | NPU | -- |
| Yolo-v7 | w8a8 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 2.528 ms | 0 - 30 MB | NPU | -- |
| Yolo-v7 | w8a8 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN | 2.755 ms | 1 - 5 MB | NPU | -- |
| Yolo-v7 | w8a8 | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 2.924 ms | 0 - 24 MB | NPU | -- |
| Yolo-v7 | w8a8 | SA8775P ADP | Qualcomm® SA8775P | QNN | 3.242 ms | 1 - 16 MB | NPU | -- |
| Yolo-v7 | w8a8 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 2.527 ms | 0 - 29 MB | NPU | -- |
| Yolo-v7 | w8a8 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 2.766 ms | 2 - 13 MB | NPU | -- |
| Yolo-v7 | w8a8 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 4.164 ms | 0 - 29 MB | NPU | -- |
| Yolo-v7 | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 1.698 ms | 0 - 37 MB | NPU | -- |
| Yolo-v7 | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 1.766 ms | 1 - 38 MB | NPU | -- |
| Yolo-v7 | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 2.742 ms | 0 - 69 MB | NPU | -- |
| Yolo-v7 | w8a8 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 1.643 ms | 0 - 28 MB | NPU | -- |
| Yolo-v7 | w8a8 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN | 1.922 ms | 1 - 30 MB | NPU | -- |
| Yolo-v7 | w8a8 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 2.822 ms | 0 - 57 MB | NPU | -- |
| Yolo-v7 | w8a8 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 3.169 ms | 1 - 1 MB | NPU | -- |
| Yolo-v7 | w8a8 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 4.498 ms | 6 - 6 MB | NPU | -- |
License
- The license for the original implementation of Yolo-v7 can be found here.
- The license for the compiled assets for on-device deployment can be found here
References
- YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors
- Source Model Implementation
Community
- Join our AI Hub Slack community to collaborate, post questions and learn more about on-device AI.
- For questions or feedback please reach out to us.
Usage and Limitations
Model may not be used for or in connection with any of the following applications:
- Accessing essential private and public services and benefits;
- Administration of justice and democratic processes;
- Assessing or recognizing the emotional state of a person;
- Biometric and biometrics-based systems, including categorization of persons based on sensitive characteristics;
- Education and vocational training;
- Employment and workers management;
- Exploitation of the vulnerabilities of persons resulting in harmful behavior;
- General purpose social scoring;
- Law enforcement;
- Management and operation of critical infrastructure;
- Migration, asylum and border control management;
- Predictive policing;
- Real-time remote biometric identification in public spaces;
- Recommender systems of social media platforms;
- Scraping of facial images (from the internet or otherwise); and/or
- Subliminal manipulation
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support