File size: 5,656 Bytes
7c53ddc | 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: other
tags:
- android
pipeline_tag: image-segmentation
---
# MaskRCNN: Optimized for Qualcomm Devices
Mask R-CNN is a machine learning model that extends Faster R-CNN to perform instance segmentation by detecting objects in an image while simultaneously generating a high-quality segmentation mask for each instance. It adds a branch for predicting segmentation masks in parallel with the existing branch for bounding box recognition.
This is based on the implementation of MaskRCNN found [here](https://github.com/pytorch/vision).
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/main/qai_hub_models/models/maskrcnn) 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
There are two ways to deploy this model on your device:
### Option 1: Download Pre-Exported Models
Below are pre-exported model assets ready for deployment.
| Runtime | Precision | Chipset | SDK Versions | Download |
|---|---|---|---|---|
| QNN_DLC | float | Universal | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/maskrcnn/releases/v0.48.0/maskrcnn-qnn_dlc-float.zip)
For more device-specific assets and performance metrics, visit **[MaskRCNN on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/maskrcnn)**.
### Option 2: Export with Custom Configurations
Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/maskrcnn) 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
This option is ideal if you need to customize the model beyond the default configuration provided here.
See our repository for [MaskRCNN on GitHub](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/maskrcnn) for usage instructions.
## Model Details
**Model Type:** Model_use_case.semantic_segmentation
**Model Stats:**
- Model checkpoint: Mask R-CNN ResNet-50 FPN V2
- Input resolution: 800x800
- Number of output classes: 91
- Number of parameters: 46.4M
- Model size (float): 177 MB
## Performance Summary
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|---|---|---|---|---|---|---
| MaskRCNNProposalGenerator | QNN_DLC | float | Snapdragon® X2 Elite | 58.538 ms | 7 - 7 MB | NPU
| MaskRCNNProposalGenerator | QNN_DLC | float | Snapdragon® X Elite | 139.069 ms | 7 - 7 MB | NPU
| MaskRCNNProposalGenerator | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 107.556 ms | 7 - 1397 MB | NPU
| MaskRCNNProposalGenerator | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 415.682 ms | 1 - 1194 MB | NPU
| MaskRCNNProposalGenerator | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 148.984 ms | 7 - 10 MB | NPU
| MaskRCNNProposalGenerator | QNN_DLC | float | Qualcomm® SA8775P | 167.04 ms | 1 - 1195 MB | NPU
| MaskRCNNProposalGenerator | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 209.409 ms | 7 - 1358 MB | NPU
| MaskRCNNProposalGenerator | QNN_DLC | float | Qualcomm® SA7255P | 415.682 ms | 1 - 1194 MB | NPU
| MaskRCNNProposalGenerator | QNN_DLC | float | Qualcomm® SA8295P | 168.867 ms | 0 - 1139 MB | NPU
| MaskRCNNProposalGenerator | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 69.508 ms | 7 - 1305 MB | NPU
| MaskRCNNProposalGenerator | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 54.824 ms | 7 - 1326 MB | NPU
| MaskRCNNROIHead | QNN_DLC | float | Snapdragon® X2 Elite | 100.247 ms | 52 - 52 MB | NPU
| MaskRCNNROIHead | QNN_DLC | float | Snapdragon® X Elite | 239.818 ms | 52 - 52 MB | NPU
| MaskRCNNROIHead | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 180.07 ms | 12 - 892 MB | NPU
| MaskRCNNROIHead | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 581.756 ms | 45 - 845 MB | NPU
| MaskRCNNROIHead | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 240.254 ms | 39 - 42 MB | NPU
| MaskRCNNROIHead | QNN_DLC | float | Qualcomm® SA8775P | 268.815 ms | 42 - 844 MB | NPU
| MaskRCNNROIHead | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 346.463 ms | 39 - 941 MB | NPU
| MaskRCNNROIHead | QNN_DLC | float | Qualcomm® SA7255P | 581.756 ms | 45 - 845 MB | NPU
| MaskRCNNROIHead | QNN_DLC | float | Qualcomm® SA8295P | 307.228 ms | 46 - 971 MB | NPU
| MaskRCNNROIHead | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 125.821 ms | 23 - 813 MB | NPU
| MaskRCNNROIHead | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 96.944 ms | 51 - 854 MB | NPU
## License
* The license for the original implementation of MaskRCNN can be found
[here](https://github.com/pytorch/vision/blob/main/LICENSE).
## References
* [Mask R-CNN](https://arxiv.org/abs/1703.06870)
* [Source Model Implementation](https://github.com/pytorch/vision)
## 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).
|