--- library_name: pytorch license: apache-2.0 tags: - real_time - android pipeline_tag: object-detection --- ![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/rtmdet/web-assets/model_demo.png) # RTMDet: Optimized for Qualcomm Devices RTMDet is a highly efficient model for real-time object detection,capable of predicting both the bounding boxes and classes of objects within an image.It is highly optimized for real-time applications, making it reliable for industrial and commercial use This is based on the implementation of RTMDet found [here](https://github.com/open-mmlab/mmdetection/tree/3.x/configs/rtmdet). 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/v0.65.0/src/qai_hub_models/models/rtmdet) 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/qualcomm/ai-hub-models/blob/v0.65.0/src/qai_hub_models/models/rtmdet) 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 [RTMDet on GitHub](https://github.com/qualcomm/ai-hub-models/blob/v0.65.0/src/qai_hub_models/models/rtmdet) for usage instructions. ## Model Details **Model Type:** Model_use_case.object_detection **Model Stats:** - Input resolution: 640x640 - Model checkpoint: RTMDet Medium - Model size (float): 105 MB - Number of parameters: 27.5M ## Performance Summary | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit |---|---|---|---|---|---|--- | RTMDet | ONNX | float | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 6.994 ms | 3 - 188 MB | NPU | RTMDet | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 9.384 ms | 2 - 183 MB | NPU | RTMDet | ONNX | float | Snapdragon® X2 Elite | 8.859 ms | 5 - 5 MB | NPU | RTMDet | ONNX | float | Snapdragon® X Elite | 16.85 ms | 51 - 51 MB | NPU | RTMDet | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 12.274 ms | 2 - 226 MB | NPU | RTMDet | ONNX | float | Snapdragon® 8 Gen 1 Mobile | 31.435 ms | 5 - 294 MB | NPU | RTMDet | ONNX | float | Qualcomm® Dragonwing™ IQ-8275 | 26.747 ms | 5 - 13 MB | NPU | RTMDet | ONNX | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 16.296 ms | 5 - 7 MB | NPU | RTMDet | ONNX | float | Qualcomm® QCS8450 | 31.435 ms | 5 - 294 MB | NPU | RTMDet | ONNX | float | Qualcomm® Dragonwing™ IQ-9075 | 23.734 ms | 5 - 13 MB | NPU | RTMDet | ONNX | float | Qualcomm® Dragonwing™ IQ-X7181 | 16.85 ms | 51 - 51 MB | NPU | RTMDet | ONNX | float | Qualcomm® Dragonwing™ Q-8750 | 9.384 ms | 2 - 183 MB | NPU | RTMDet | ONNX | w8a16 | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 5.93 ms | 0 - 314 MB | NPU | RTMDet | ONNX | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 7.169 ms | 1 - 291 MB | NPU | RTMDet | ONNX | w8a16 | Snapdragon® X2 Elite | 6.616 ms | 2 - 2 MB | NPU | RTMDet | ONNX | w8a16 | Snapdragon® X Elite | 15.281 ms | 27 - 27 MB | NPU | RTMDet | ONNX | w8a16 | Snapdragon® 8 Gen 3 Mobile | 12.292 ms | 3 - 349 MB | NPU | RTMDet | ONNX | w8a16 | Snapdragon® 8 Gen 1 Mobile | 19.061 ms | 3 - 350 MB | NPU | RTMDet | ONNX | w8a16 | Qualcomm® Dragonwing™ QCS6490 | 53.185 ms | 23 - 26 MB | NPU | RTMDet | ONNX | w8a16 | Qualcomm® Dragonwing™ IQ-8275 | 13.841 ms | 2 - 6 MB | NPU | RTMDet | ONNX | w8a16 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 14.877 ms | 0 - 33 MB | NPU | RTMDet | ONNX | w8a16 | Qualcomm® QCS8450 | 19.061 ms | 3 - 350 MB | NPU | RTMDet | ONNX | w8a16 | Qualcomm® Dragonwing™ IQ-9075 | 15.606 ms | 2 - 6 MB | NPU | RTMDet | ONNX | w8a16 | Qualcomm® Dragonwing™ IQ-X7181 | 15.281 ms | 27 - 27 MB | NPU | RTMDet | ONNX | w8a16 | Qualcomm® Dragonwing™ Q-6690 | 138.785 ms | 23 - 331 MB | NPU | RTMDet | ONNX | w8a16 | Qualcomm® Dragonwing™ Q-7790 | 22.757 ms | 25 - 318 MB | NPU | RTMDet | ONNX | w8a16 | Qualcomm® Dragonwing™ Q-8750 | 7.169 ms | 1 - 291 MB | NPU | RTMDet | ONNX | w8a16 | Snapdragon® 7 Gen 4 Mobile | 22.757 ms | 25 - 318 MB | NPU | RTMDet | TFLITE | float | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 6.517 ms | 8 - 216 MB | NPU | RTMDet | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 9.164 ms | 0 - 203 MB | NPU | RTMDet | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 11.907 ms | 9 - 287 MB | NPU | RTMDet | TFLITE | float | Snapdragon® 8 Gen 1 Mobile | 37.571 ms | 10 - 352 MB | NPU | RTMDet | TFLITE | float | Qualcomm® Dragonwing™ IQ-8275 | 26.996 ms | 9 - 81 MB | NPU | RTMDet | TFLITE | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 15.676 ms | 9 - 12 MB | NPU | RTMDet | TFLITE | float | Qualcomm® SA8775P | 23.907 ms | 9 - 212 MB | NPU | RTMDet | TFLITE | float | Qualcomm® SA8650P | 23.907 ms | 9 - 212 MB | NPU | RTMDet | TFLITE | float | Qualcomm® SA8255P | 23.907 ms | 9 - 212 MB | NPU | RTMDet | TFLITE | float | Qualcomm® QCS8450 | 37.571 ms | 10 - 352 MB | NPU | RTMDet | TFLITE | float | Qualcomm® Dragonwing™ IQ-9075 | 23.971 ms | 8 - 79 MB | NPU | RTMDet | TFLITE | float | Qualcomm® Dragonwing™ Q-8750 | 9.164 ms | 0 - 203 MB | NPU | RTMDet | TFLITE | float | Qualcomm® SA7255P | 84.857 ms | 10 - 211 MB | NPU | RTMDet | TFLITE | float | Qualcomm® SA8295P | 32.632 ms | 10 - 275 MB | NPU ## License * The license for the original implementation of RTMDet can be found [here](https://github.com/open-mmlab/mmdetection/blob/3.x/LICENSE). ## References * [RTMDet: An Empirical Study of Designing Real-Time Object Detectors](https://github.com/open-mmlab/mmdetection/blob/3.x/README.md) * [Source Model Implementation](https://github.com/open-mmlab/mmdetection/tree/3.x/configs/rtmdet) ## 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).