v0.48.0
Browse filesSee https://github.com/qualcomm/ai-hub-models/releases/v0.48.0 for changelog.
README.md
CHANGED
|
@@ -15,7 +15,7 @@ pipeline_tag: image-segmentation
|
|
| 15 |
Transformer based encoder-decoder where prompts specify what to segment in an image thereby allowing segmentation without the need for additional training. The image encoder generates embeddings and the lightweight decoder operates on the embeddings for point and mask based image segmentation.
|
| 16 |
|
| 17 |
This is based on the implementation of MobileSam found [here](https://github.com/facebookresearch/segment-anything).
|
| 18 |
-
This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/
|
| 19 |
|
| 20 |
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.
|
| 21 |
|
|
@@ -28,23 +28,23 @@ Below are pre-exported model assets ready for deployment.
|
|
| 28 |
|
| 29 |
| Runtime | Precision | Chipset | SDK Versions | Download |
|
| 30 |
|---|---|---|---|---|
|
| 31 |
-
| ONNX | float | Universal | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/mobilesam/releases/v0.
|
| 32 |
-
| QNN_DLC | float | Universal | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/mobilesam/releases/v0.
|
| 33 |
-
| TFLITE | float | Universal | QAIRT 2.43, TFLite 2.17.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/mobilesam/releases/v0.
|
| 34 |
|
| 35 |
For more device-specific assets and performance metrics, visit **[MobileSam on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/mobilesam)**.
|
| 36 |
|
| 37 |
|
| 38 |
### Option 2: Export with Custom Configurations
|
| 39 |
|
| 40 |
-
Use the [Qualcomm® AI Hub Models](https://github.com/
|
| 41 |
- Custom weights (e.g., fine-tuned checkpoints)
|
| 42 |
- Custom input shapes
|
| 43 |
- Target device and runtime configurations
|
| 44 |
|
| 45 |
This option is ideal if you need to customize the model beyond the default configuration provided here.
|
| 46 |
|
| 47 |
-
See our repository for [MobileSam on GitHub](https://github.com/
|
| 48 |
|
| 49 |
## Model Details
|
| 50 |
|
|
@@ -61,58 +61,58 @@ See our repository for [MobileSam on GitHub](https://github.com/quic/ai-hub-mode
|
|
| 61 |
## Performance Summary
|
| 62 |
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|
| 63 |
|---|---|---|---|---|---|---
|
| 64 |
-
| MobileSAMDecoder | ONNX | float | Snapdragon®
|
| 65 |
-
| MobileSAMDecoder | ONNX | float | Snapdragon®
|
| 66 |
-
| MobileSAMDecoder | ONNX | float |
|
| 67 |
-
| MobileSAMDecoder | ONNX | float |
|
| 68 |
-
| MobileSAMDecoder | ONNX | float | Snapdragon® 8 Elite
|
| 69 |
-
| MobileSAMDecoder | ONNX | float | Snapdragon®
|
| 70 |
-
| MobileSAMDecoder | QNN_DLC | float | Snapdragon®
|
| 71 |
-
| MobileSAMDecoder | QNN_DLC | float | Snapdragon®
|
| 72 |
-
| MobileSAMDecoder | QNN_DLC | float |
|
| 73 |
-
| MobileSAMDecoder | QNN_DLC | float | Qualcomm®
|
| 74 |
-
| MobileSAMDecoder | QNN_DLC | float | Qualcomm®
|
| 75 |
-
| MobileSAMDecoder | QNN_DLC | float | Qualcomm®
|
| 76 |
-
| MobileSAMDecoder | QNN_DLC | float | Qualcomm®
|
| 77 |
-
| MobileSAMDecoder | QNN_DLC | float | Qualcomm®
|
| 78 |
-
| MobileSAMDecoder | QNN_DLC | float |
|
| 79 |
-
| MobileSAMDecoder | QNN_DLC | float | Snapdragon® 8 Elite
|
| 80 |
-
| MobileSAMDecoder | QNN_DLC | float | Snapdragon®
|
| 81 |
-
| MobileSAMDecoder | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 3.
|
| 82 |
-
| MobileSAMDecoder | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 13.
|
| 83 |
-
| MobileSAMDecoder | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 5.
|
| 84 |
-
| MobileSAMDecoder | TFLITE | float | Qualcomm® SA8775P | 6.
|
| 85 |
-
| MobileSAMDecoder | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 9.
|
| 86 |
-
| MobileSAMDecoder | TFLITE | float | Qualcomm® SA7255P | 13.
|
| 87 |
-
| MobileSAMDecoder | TFLITE | float | Qualcomm® SA8295P | 8.
|
| 88 |
-
| MobileSAMDecoder | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 2.
|
| 89 |
-
| MobileSAMDecoder | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 2.
|
| 90 |
-
| MobileSAMEncoder | ONNX | float | Snapdragon®
|
| 91 |
-
| MobileSAMEncoder | ONNX | float | Snapdragon®
|
| 92 |
-
| MobileSAMEncoder | ONNX | float |
|
| 93 |
-
| MobileSAMEncoder | ONNX | float |
|
| 94 |
-
| MobileSAMEncoder | ONNX | float | Snapdragon® 8 Elite
|
| 95 |
-
| MobileSAMEncoder | ONNX | float | Snapdragon®
|
| 96 |
-
| MobileSAMEncoder | QNN_DLC | float | Snapdragon®
|
| 97 |
-
| MobileSAMEncoder | QNN_DLC | float | Snapdragon®
|
| 98 |
-
| MobileSAMEncoder | QNN_DLC | float |
|
| 99 |
-
| MobileSAMEncoder | QNN_DLC | float | Qualcomm®
|
| 100 |
-
| MobileSAMEncoder | QNN_DLC | float | Qualcomm®
|
| 101 |
-
| MobileSAMEncoder | QNN_DLC | float | Qualcomm®
|
| 102 |
-
| MobileSAMEncoder | QNN_DLC | float | Qualcomm®
|
| 103 |
-
| MobileSAMEncoder | QNN_DLC | float | Qualcomm®
|
| 104 |
-
| MobileSAMEncoder | QNN_DLC | float |
|
| 105 |
-
| MobileSAMEncoder | QNN_DLC | float | Snapdragon® 8 Elite
|
| 106 |
-
| MobileSAMEncoder | QNN_DLC | float | Snapdragon®
|
| 107 |
-
| MobileSAMEncoder | TFLITE | float | Snapdragon® 8 Gen 3 Mobile |
|
| 108 |
-
| MobileSAMEncoder | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 295.
|
| 109 |
-
| MobileSAMEncoder | TFLITE | float | Qualcomm® QCS8550 (Proxy) |
|
| 110 |
-
| MobileSAMEncoder | TFLITE | float | Qualcomm® SA8775P |
|
| 111 |
-
| MobileSAMEncoder | TFLITE | float | Qualcomm® QCS8450 (Proxy) |
|
| 112 |
-
| MobileSAMEncoder | TFLITE | float | Qualcomm® SA7255P | 295.
|
| 113 |
-
| MobileSAMEncoder | TFLITE | float | Qualcomm® SA8295P | 502.
|
| 114 |
-
| MobileSAMEncoder | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile |
|
| 115 |
-
| MobileSAMEncoder | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 78.
|
| 116 |
|
| 117 |
## License
|
| 118 |
* The license for the original implementation of MobileSam can be found
|
|
|
|
| 15 |
Transformer based encoder-decoder where prompts specify what to segment in an image thereby allowing segmentation without the need for additional training. The image encoder generates embeddings and the lightweight decoder operates on the embeddings for point and mask based image segmentation.
|
| 16 |
|
| 17 |
This is based on the implementation of MobileSam found [here](https://github.com/facebookresearch/segment-anything).
|
| 18 |
+
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/mobilesam) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
|
| 19 |
|
| 20 |
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.
|
| 21 |
|
|
|
|
| 28 |
|
| 29 |
| Runtime | Precision | Chipset | SDK Versions | Download |
|
| 30 |
|---|---|---|---|---|
|
| 31 |
+
| ONNX | float | Universal | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/mobilesam/releases/v0.48.0/mobilesam-onnx-float.zip)
|
| 32 |
+
| QNN_DLC | float | Universal | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/mobilesam/releases/v0.48.0/mobilesam-qnn_dlc-float.zip)
|
| 33 |
+
| TFLITE | float | Universal | QAIRT 2.43, TFLite 2.17.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/mobilesam/releases/v0.48.0/mobilesam-tflite-float.zip)
|
| 34 |
|
| 35 |
For more device-specific assets and performance metrics, visit **[MobileSam on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/mobilesam)**.
|
| 36 |
|
| 37 |
|
| 38 |
### Option 2: Export with Custom Configurations
|
| 39 |
|
| 40 |
+
Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/mobilesam) Python library to compile and export the model with your own:
|
| 41 |
- Custom weights (e.g., fine-tuned checkpoints)
|
| 42 |
- Custom input shapes
|
| 43 |
- Target device and runtime configurations
|
| 44 |
|
| 45 |
This option is ideal if you need to customize the model beyond the default configuration provided here.
|
| 46 |
|
| 47 |
+
See our repository for [MobileSam on GitHub](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/mobilesam) for usage instructions.
|
| 48 |
|
| 49 |
## Model Details
|
| 50 |
|
|
|
|
| 61 |
## Performance Summary
|
| 62 |
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|
| 63 |
|---|---|---|---|---|---|---
|
| 64 |
+
| MobileSAMDecoder | ONNX | float | Snapdragon® X2 Elite | 2.953 ms | 11 - 11 MB | NPU
|
| 65 |
+
| MobileSAMDecoder | ONNX | float | Snapdragon® X Elite | 7.521 ms | 11 - 11 MB | NPU
|
| 66 |
+
| MobileSAMDecoder | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 4.892 ms | 4 - 270 MB | NPU
|
| 67 |
+
| MobileSAMDecoder | ONNX | float | Qualcomm® QCS8550 (Proxy) | 6.988 ms | 0 - 15 MB | NPU
|
| 68 |
+
| MobileSAMDecoder | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 3.641 ms | 1 - 223 MB | NPU
|
| 69 |
+
| MobileSAMDecoder | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 2.886 ms | 0 - 232 MB | NPU
|
| 70 |
+
| MobileSAMDecoder | QNN_DLC | float | Snapdragon® X2 Elite | 2.609 ms | 4 - 4 MB | NPU
|
| 71 |
+
| MobileSAMDecoder | QNN_DLC | float | Snapdragon® X Elite | 5.309 ms | 4 - 4 MB | NPU
|
| 72 |
+
| MobileSAMDecoder | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 3.364 ms | 4 - 230 MB | NPU
|
| 73 |
+
| MobileSAMDecoder | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 11.701 ms | 0 - 199 MB | NPU
|
| 74 |
+
| MobileSAMDecoder | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 4.844 ms | 4 - 5 MB | NPU
|
| 75 |
+
| MobileSAMDecoder | QNN_DLC | float | Qualcomm® SA8775P | 5.875 ms | 0 - 200 MB | NPU
|
| 76 |
+
| MobileSAMDecoder | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 8.445 ms | 0 - 220 MB | NPU
|
| 77 |
+
| MobileSAMDecoder | QNN_DLC | float | Qualcomm® SA7255P | 11.701 ms | 0 - 199 MB | NPU
|
| 78 |
+
| MobileSAMDecoder | QNN_DLC | float | Qualcomm® SA8295P | 7.169 ms | 0 - 193 MB | NPU
|
| 79 |
+
| MobileSAMDecoder | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 2.593 ms | 0 - 200 MB | NPU
|
| 80 |
+
| MobileSAMDecoder | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 2.151 ms | 2 - 201 MB | NPU
|
| 81 |
+
| MobileSAMDecoder | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 3.877 ms | 0 - 235 MB | NPU
|
| 82 |
+
| MobileSAMDecoder | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 13.033 ms | 0 - 202 MB | NPU
|
| 83 |
+
| MobileSAMDecoder | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 5.605 ms | 0 - 18 MB | NPU
|
| 84 |
+
| MobileSAMDecoder | TFLITE | float | Qualcomm® SA8775P | 6.698 ms | 0 - 203 MB | NPU
|
| 85 |
+
| MobileSAMDecoder | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 9.541 ms | 0 - 225 MB | NPU
|
| 86 |
+
| MobileSAMDecoder | TFLITE | float | Qualcomm® SA7255P | 13.033 ms | 0 - 202 MB | NPU
|
| 87 |
+
| MobileSAMDecoder | TFLITE | float | Qualcomm® SA8295P | 8.047 ms | 0 - 201 MB | NPU
|
| 88 |
+
| MobileSAMDecoder | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 2.958 ms | 0 - 207 MB | NPU
|
| 89 |
+
| MobileSAMDecoder | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 2.352 ms | 0 - 207 MB | NPU
|
| 90 |
+
| MobileSAMEncoder | ONNX | float | Snapdragon® X2 Elite | 94.358 ms | 14 - 14 MB | NPU
|
| 91 |
+
| MobileSAMEncoder | ONNX | float | Snapdragon® X Elite | 173.085 ms | 12 - 12 MB | NPU
|
| 92 |
+
| MobileSAMEncoder | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 125.467 ms | 20 - 3025 MB | NPU
|
| 93 |
+
| MobileSAMEncoder | ONNX | float | Qualcomm® QCS8550 (Proxy) | 175.267 ms | 12 - 32 MB | NPU
|
| 94 |
+
| MobileSAMEncoder | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 100.697 ms | 9 - 1389 MB | NPU
|
| 95 |
+
| MobileSAMEncoder | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 88.418 ms | 20 - 1523 MB | NPU
|
| 96 |
+
| MobileSAMEncoder | QNN_DLC | float | Snapdragon® X2 Elite | 51.746 ms | 12 - 12 MB | NPU
|
| 97 |
+
| MobileSAMEncoder | QNN_DLC | float | Snapdragon® X Elite | 100.135 ms | 12 - 12 MB | NPU
|
| 98 |
+
| MobileSAMEncoder | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 66.42 ms | 12 - 1635 MB | NPU
|
| 99 |
+
| MobileSAMEncoder | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 209.122 ms | 0 - 1209 MB | NPU
|
| 100 |
+
| MobileSAMEncoder | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 96.617 ms | 12 - 16 MB | NPU
|
| 101 |
+
| MobileSAMEncoder | QNN_DLC | float | Qualcomm® SA8775P | 103.074 ms | 0 - 1209 MB | NPU
|
| 102 |
+
| MobileSAMEncoder | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 470.205 ms | 12 - 1731 MB | NPU
|
| 103 |
+
| MobileSAMEncoder | QNN_DLC | float | Qualcomm® SA7255P | 209.122 ms | 0 - 1209 MB | NPU
|
| 104 |
+
| MobileSAMEncoder | QNN_DLC | float | Qualcomm® SA8295P | 474.048 ms | 0 - 1249 MB | NPU
|
| 105 |
+
| MobileSAMEncoder | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 49.865 ms | 12 - 1217 MB | NPU
|
| 106 |
+
| MobileSAMEncoder | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 48.392 ms | 8 - 1319 MB | NPU
|
| 107 |
+
| MobileSAMEncoder | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 126.457 ms | 0 - 1496 MB | NPU
|
| 108 |
+
| MobileSAMEncoder | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 295.258 ms | 4 - 1269 MB | NPU
|
| 109 |
+
| MobileSAMEncoder | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 176.603 ms | 0 - 3 MB | NPU
|
| 110 |
+
| MobileSAMEncoder | TFLITE | float | Qualcomm® SA8775P | 179.751 ms | 0 - 1265 MB | NPU
|
| 111 |
+
| MobileSAMEncoder | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 590.208 ms | 0 - 3175 MB | NPU
|
| 112 |
+
| MobileSAMEncoder | TFLITE | float | Qualcomm® SA7255P | 295.258 ms | 4 - 1269 MB | NPU
|
| 113 |
+
| MobileSAMEncoder | TFLITE | float | Qualcomm® SA8295P | 502.789 ms | 0 - 1888 MB | NPU
|
| 114 |
+
| MobileSAMEncoder | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 97.396 ms | 4 - 1267 MB | NPU
|
| 115 |
+
| MobileSAMEncoder | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 78.068 ms | 4 - 1348 MB | NPU
|
| 116 |
|
| 117 |
## License
|
| 118 |
* The license for the original implementation of MobileSam can be found
|