--- library_name: pytorch license: other tags: - bu_auto - real_time - android pipeline_tag: object-detection --- ![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/yolor/web-assets/model_demo.png) # Yolo-R: Optimized for Qualcomm Devices YoloR is a machine learning model that predicts bounding boxes and classes of objects in an image. This is based on the implementation of Yolo-R found [here](https://github.com/WongKinYiu/yolor.git). This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/yolor) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary). 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. ## Getting Started Due to licensing restrictions, we cannot distribute pre-exported model assets for this model. Use the [Qualcomm® AI Hub Models](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/yolor) Python library to compile and export the model with your own: - Custom weights (e.g., fine-tuned checkpoints) - Custom input shapes - Target device and runtime configurations See our repository for [Yolo-R on GitHub](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/yolor) for usage instructions. ## Model Details **Model Type:** Model_use_case.object_detection **Model Stats:** - Model checkpoint: yolor_p6 - Input resolution: 640x640 - Number of parameters: 4.68M - Model size (float): 17.9 MB ## Performance Summary | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit |---|---|---|---|---|---|--- | Yolo-R | ONNX | float | Snapdragon® X Elite | 59.708 ms | 75 - 75 MB | NPU | Yolo-R | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 46.21 ms | 2 - 291 MB | NPU | Yolo-R | ONNX | float | Qualcomm® QCS8550 (Proxy) | 63.877 ms | 5 - 8 MB | NPU | Yolo-R | ONNX | float | Qualcomm® QCS9075 | 81.653 ms | 5 - 12 MB | NPU | Yolo-R | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 32.462 ms | 1 - 173 MB | NPU | Yolo-R | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 30.32 ms | 0 - 245 MB | NPU | Yolo-R | ONNX | w8a16 | Snapdragon® X Elite | 31.604 ms | 41 - 41 MB | NPU | Yolo-R | ONNX | w8a16 | Snapdragon® 8 Gen 3 Mobile | 22.018 ms | 1 - 397 MB | NPU | Yolo-R | ONNX | w8a16 | Qualcomm® QCS6490 | 2301.178 ms | 124 - 142 MB | CPU | Yolo-R | ONNX | w8a16 | Qualcomm® QCS8550 (Proxy) | 37.456 ms | 0 - 424 MB | NPU | Yolo-R | ONNX | w8a16 | Qualcomm® QCS9075 | 32.023 ms | 1 - 6 MB | NPU | Yolo-R | ONNX | w8a16 | Qualcomm® QCM6690 | 1155.802 ms | 36 - 48 MB | CPU | Yolo-R | ONNX | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 17.63 ms | 1 - 292 MB | NPU | Yolo-R | ONNX | w8a16 | Snapdragon® 7 Gen 4 Mobile | 1089.713 ms | 124 - 137 MB | CPU | Yolo-R | ONNX | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | 18.688 ms | 0 - 368 MB | NPU | Yolo-R | QNN_DLC | float | Snapdragon® X Elite | 29.858 ms | 5 - 5 MB | NPU | Yolo-R | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 22.895 ms | 5 - 344 MB | NPU | Yolo-R | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 100.474 ms | 1 - 273 MB | NPU | Yolo-R | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 30.237 ms | 5 - 7 MB | NPU | Yolo-R | QNN_DLC | float | Qualcomm® SA8775P | 35.561 ms | 1 - 276 MB | NPU | Yolo-R | QNN_DLC | float | Qualcomm® QCS9075 | 37.297 ms | 5 - 11 MB | NPU | Yolo-R | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 49.525 ms | 4 - 379 MB | NPU | Yolo-R | QNN_DLC | float | Qualcomm® SA7255P | 100.474 ms | 1 - 273 MB | NPU | Yolo-R | QNN_DLC | float | Qualcomm® SA8295P | 43.831 ms | 0 - 321 MB | NPU | Yolo-R | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 18.211 ms | 5 - 237 MB | NPU | Yolo-R | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 14.214 ms | 5 - 313 MB | NPU | Yolo-R | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 38.894 ms | 0 - 449 MB | NPU | Yolo-R | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 135.572 ms | 1 - 323 MB | NPU | Yolo-R | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 58.257 ms | 1 - 3 MB | NPU | Yolo-R | TFLITE | float | Qualcomm® SA8775P | 61.1 ms | 1 - 338 MB | NPU | Yolo-R | TFLITE | float | Qualcomm® QCS9075 | 50.135 ms | 1 - 86 MB | NPU | Yolo-R | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 87.963 ms | 1 - 470 MB | NPU | Yolo-R | TFLITE | float | Qualcomm® SA7255P | 135.572 ms | 1 - 323 MB | NPU | Yolo-R | TFLITE | float | Qualcomm® SA8295P | 73.377 ms | 1 - 356 MB | NPU | Yolo-R | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 31.364 ms | 0 - 281 MB | NPU | Yolo-R | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 25.844 ms | 1 - 344 MB | NPU ## License * The license for the original implementation of Yolo-R can be found [here](https://github.com/WongKinYiu/yolor/blob/main/LICENSE). ## References * [You Only Learn One Representation: Unified Network for Multiple Tasks](https://arxiv.org/abs/2105.04206) * [Source Model Implementation](https://github.com/WongKinYiu/yolor.git) ## Community * Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI. * For questions or feedback please [reach out to us](mailto:ai-hub-support@qti.qualcomm.com).