File size: 4,640 Bytes
0be84dd 12119a4 0be84dd 12119a4 0be84dd 12119a4 0be84dd 84612ed 0be84dd 12119a4 0be84dd 12119a4 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 |
---
library_name: pytorch
license: agpl-3.0
tags:
- real_time
- android
pipeline_tag: object-detection
---
# Yolo-v5: Optimized for Mobile Deployment
## Real-time object detection optimized for mobile and edge
YoloV5 is a machine learning model that predicts bounding boxes and classes of objects in an image.
This model is an implementation of Yolo-v5 found [here](https://github.com/ultralytics/yolov5).
More details on model performance across various devices, can be found [here](https://aihub.qualcomm.com/models/yolov5).
### Model Details
- **Model Type:** Object detection
- **Model Stats:**
- Model checkpoint: YoloV5-M
- Input resolution: 640x640
- Number of parameters: 21.2M
- Model size: 81.1 MB
| Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
|---|---|---|---|---|---|---|---|---|
| Yolo-v5 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 23.747 ms | 6 - 38 MB | FP16 | NPU | -- |
| Yolo-v5 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 21.883 ms | 6 - 8 MB | FP16 | NPU | -- |
| Yolo-v5 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 27.816 ms | 1 - 119 MB | FP16 | NPU | -- |
| Yolo-v5 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 18.188 ms | 5 - 104 MB | FP16 | NPU | -- |
| Yolo-v5 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 16.767 ms | 5 - 25 MB | FP16 | NPU | -- |
| Yolo-v5 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 21.418 ms | 7 - 134 MB | FP16 | NPU | -- |
| Yolo-v5 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 16.687 ms | 5 - 82 MB | FP16 | NPU | -- |
| Yolo-v5 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 15.715 ms | 5 - 128 MB | FP16 | NPU | -- |
| Yolo-v5 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 20.519 ms | 5 - 120 MB | FP16 | NPU | -- |
| Yolo-v5 | QCS8275 (Proxy) | QCS8275 Proxy | TFLITE | 370.263 ms | 1 - 74 MB | FP16 | NPU | -- |
| Yolo-v5 | QCS8275 (Proxy) | QCS8275 Proxy | QNN | 364.102 ms | 1 - 10 MB | FP16 | NPU | -- |
| Yolo-v5 | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 23.754 ms | 6 - 39 MB | FP16 | NPU | -- |
| Yolo-v5 | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 22.141 ms | 5 - 7 MB | FP16 | NPU | -- |
| Yolo-v5 | QCS9075 (Proxy) | QCS9075 Proxy | TFLITE | 34.771 ms | 0 - 74 MB | FP16 | NPU | -- |
| Yolo-v5 | QCS9075 (Proxy) | QCS9075 Proxy | QNN | 30.88 ms | 1 - 11 MB | FP16 | NPU | -- |
| Yolo-v5 | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 34.864 ms | 6 - 87 MB | FP16 | NPU | -- |
| Yolo-v5 | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 42.293 ms | 5 - 44 MB | FP16 | NPU | -- |
| Yolo-v5 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 21.469 ms | 5 - 5 MB | FP16 | NPU | -- |
| Yolo-v5 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 28.022 ms | 39 - 39 MB | FP16 | NPU | -- |
## License
* The license for the original implementation of Yolo-v5 can be found
[here](https://github.com/ultralytics/yolov5?tab=AGPL-3.0-1-ov-file#readme).
* The license for the compiled assets for on-device deployment can be found [here](https://github.com/ultralytics/yolov5?tab=AGPL-3.0-1-ov-file#readme)
## References
* [Source Model Implementation](https://github.com/ultralytics/yolov5)
## Community
* Join [our AI Hub Slack community](https://qualcomm-ai-hub.slack.com/join/shared_invite/zt-2d5zsmas3-Sj0Q9TzslueCjS31eXG2UA#/shared-invite/email) to collaborate, post questions and learn more about on-device AI.
* For questions or feedback please [reach out to us](mailto:[email protected]).
## 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
|