v0.50.0
Browse filesSee https://github.com/qualcomm/ai-hub-models/releases/v0.50.0 for changelog.
- README.md +77 -77
- release_assets.json +1 -0
README.md
CHANGED
|
@@ -14,7 +14,7 @@ pipeline_tag: video-object-tracking
|
|
| 14 |
Track-Anything is a video based machine learning model to track an object in a video.
|
| 15 |
|
| 16 |
This is based on the implementation of Track-Anything found [here](https://github.com/gaomingqi/Track-Anything).
|
| 17 |
-
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/
|
| 18 |
|
| 19 |
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.
|
| 20 |
|
|
@@ -27,22 +27,22 @@ Below are pre-exported model assets ready for deployment.
|
|
| 27 |
|
| 28 |
| Runtime | Precision | Chipset | SDK Versions | Download |
|
| 29 |
|---|---|---|---|---|
|
| 30 |
-
| 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/track_anything/releases/v0.
|
| 31 |
-
| 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/track_anything/releases/v0.
|
| 32 |
|
| 33 |
For more device-specific assets and performance metrics, visit **[Track-Anything on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/track_anything)**.
|
| 34 |
|
| 35 |
|
| 36 |
### Option 2: Export with Custom Configurations
|
| 37 |
|
| 38 |
-
Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/
|
| 39 |
- Custom weights (e.g., fine-tuned checkpoints)
|
| 40 |
- Custom input shapes
|
| 41 |
- Target device and runtime configurations
|
| 42 |
|
| 43 |
This option is ideal if you need to customize the model beyond the default configuration provided here.
|
| 44 |
|
| 45 |
-
See our repository for [Track-Anything on GitHub](https://github.com/qualcomm/ai-hub-models/
|
| 46 |
|
| 47 |
## Model Details
|
| 48 |
|
|
@@ -51,82 +51,82 @@ See our repository for [Track-Anything on GitHub](https://github.com/qualcomm/ai
|
|
| 51 |
**Model Stats:**
|
| 52 |
- Model checkpoint: xmem
|
| 53 |
- Input resolution: 320x568
|
| 54 |
-
- Number of parameters (
|
| 55 |
-
- Model size (
|
| 56 |
-
- Number of parameters (
|
| 57 |
-
- Model size (
|
| 58 |
-
- Number of parameters (
|
| 59 |
-
- Model size (
|
| 60 |
-
- Number of parameters (
|
| 61 |
-
- Model size (
|
| 62 |
|
| 63 |
## Performance Summary
|
| 64 |
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|
| 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 |
-
|
|
| 95 |
-
|
|
| 96 |
-
|
|
| 97 |
-
|
|
| 98 |
-
|
|
| 99 |
-
|
|
| 100 |
-
|
|
| 101 |
-
|
|
| 102 |
-
|
|
| 103 |
-
|
|
| 104 |
-
|
|
| 105 |
-
|
|
| 106 |
-
|
|
| 107 |
-
|
|
| 108 |
-
|
|
| 109 |
-
|
|
| 110 |
-
|
|
| 111 |
-
|
|
| 112 |
-
|
|
| 113 |
-
|
|
| 114 |
-
|
|
| 115 |
-
|
|
| 116 |
-
|
|
| 117 |
-
|
|
| 118 |
-
|
|
| 119 |
-
|
|
| 120 |
-
|
|
| 121 |
-
|
|
| 122 |
-
|
|
| 123 |
-
|
|
| 124 |
-
|
|
| 125 |
-
|
|
| 126 |
-
|
|
| 127 |
-
|
|
| 128 |
-
|
|
| 129 |
-
|
|
| 130 |
|
| 131 |
## License
|
| 132 |
* The license for the original implementation of Track-Anything can be found
|
|
|
|
| 14 |
Track-Anything is a video based machine learning model to track an object in a video.
|
| 15 |
|
| 16 |
This is based on the implementation of Track-Anything found [here](https://github.com/gaomingqi/Track-Anything).
|
| 17 |
+
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/track_anything) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
|
| 18 |
|
| 19 |
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.
|
| 20 |
|
|
|
|
| 27 |
|
| 28 |
| Runtime | Precision | Chipset | SDK Versions | Download |
|
| 29 |
|---|---|---|---|---|
|
| 30 |
+
| 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/track_anything/releases/v0.50.0/track_anything-onnx-float.zip)
|
| 31 |
+
| 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/track_anything/releases/v0.50.0/track_anything-tflite-float.zip)
|
| 32 |
|
| 33 |
For more device-specific assets and performance metrics, visit **[Track-Anything on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/track_anything)**.
|
| 34 |
|
| 35 |
|
| 36 |
### Option 2: Export with Custom Configurations
|
| 37 |
|
| 38 |
+
Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/track_anything) Python library to compile and export the model with your own:
|
| 39 |
- Custom weights (e.g., fine-tuned checkpoints)
|
| 40 |
- Custom input shapes
|
| 41 |
- Target device and runtime configurations
|
| 42 |
|
| 43 |
This option is ideal if you need to customize the model beyond the default configuration provided here.
|
| 44 |
|
| 45 |
+
See our repository for [Track-Anything on GitHub](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/track_anything) for usage instructions.
|
| 46 |
|
| 47 |
## Model Details
|
| 48 |
|
|
|
|
| 51 |
**Model Stats:**
|
| 52 |
- Model checkpoint: xmem
|
| 53 |
- Input resolution: 320x568
|
| 54 |
+
- Number of parameters (encode_key_with_shrinkage): 9.72M
|
| 55 |
+
- Model size (encode_key_with_shrinkage) (float): 37.1 MB
|
| 56 |
+
- Number of parameters (encode_value): 23.3M
|
| 57 |
+
- Model size (encode_value) (float): 88.8 MB
|
| 58 |
+
- Number of parameters (encode_key_without_shrinkage): 9.71M
|
| 59 |
+
- Model size (encode_key_without_shrinkage) (float): 37.1 MB
|
| 60 |
+
- Number of parameters (segment): 30.1M
|
| 61 |
+
- Model size (segment) (float): 115 MB
|
| 62 |
|
| 63 |
## Performance Summary
|
| 64 |
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|
| 65 |
|---|---|---|---|---|---|---
|
| 66 |
+
| encode_key_with_shrinkage | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 2.384 ms | 1 - 177 MB | NPU
|
| 67 |
+
| encode_key_with_shrinkage | ONNX | float | Snapdragon® X2 Elite | 2.813 ms | 18 - 18 MB | NPU
|
| 68 |
+
| encode_key_with_shrinkage | ONNX | float | Snapdragon® X Elite | 5.608 ms | 19 - 19 MB | NPU
|
| 69 |
+
| encode_key_with_shrinkage | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 3.664 ms | 0 - 202 MB | NPU
|
| 70 |
+
| encode_key_with_shrinkage | ONNX | float | Qualcomm® QCS8550 (Proxy) | 5.166 ms | 0 - 21 MB | NPU
|
| 71 |
+
| encode_key_with_shrinkage | ONNX | float | Qualcomm® QCS9075 | 9.464 ms | 10 - 13 MB | NPU
|
| 72 |
+
| encode_key_with_shrinkage | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 2.948 ms | 1 - 166 MB | NPU
|
| 73 |
+
| encode_key_with_shrinkage | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 2.441 ms | 0 - 191 MB | NPU
|
| 74 |
+
| encode_key_with_shrinkage | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 4.085 ms | 1 - 213 MB | NPU
|
| 75 |
+
| encode_key_with_shrinkage | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 31.167 ms | 3 - 180 MB | NPU
|
| 76 |
+
| encode_key_with_shrinkage | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 5.935 ms | 3 - 6 MB | NPU
|
| 77 |
+
| encode_key_with_shrinkage | TFLITE | float | Qualcomm® SA8775P | 9.553 ms | 0 - 178 MB | NPU
|
| 78 |
+
| encode_key_with_shrinkage | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 11.742 ms | 3 - 197 MB | NPU
|
| 79 |
+
| encode_key_with_shrinkage | TFLITE | float | Qualcomm® SA7255P | 31.167 ms | 3 - 180 MB | NPU
|
| 80 |
+
| encode_key_with_shrinkage | TFLITE | float | Qualcomm® SA8295P | 10.378 ms | 3 - 165 MB | NPU
|
| 81 |
+
| encode_key_with_shrinkage | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 3.289 ms | 0 - 180 MB | NPU
|
| 82 |
+
| encode_key_without_shrinkage | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 3.494 ms | 1 - 216 MB | NPU
|
| 83 |
+
| encode_key_without_shrinkage | ONNX | float | Snapdragon® X2 Elite | 3.947 ms | 65 - 65 MB | NPU
|
| 84 |
+
| encode_key_without_shrinkage | ONNX | float | Snapdragon® X Elite | 6.955 ms | 66 - 66 MB | NPU
|
| 85 |
+
| encode_key_without_shrinkage | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 5.056 ms | 63 - 271 MB | NPU
|
| 86 |
+
| encode_key_without_shrinkage | ONNX | float | Qualcomm® QCS8550 (Proxy) | 6.835 ms | 58 - 77 MB | NPU
|
| 87 |
+
| encode_key_without_shrinkage | ONNX | float | Qualcomm® QCS9075 | 11.477 ms | 62 - 64 MB | NPU
|
| 88 |
+
| encode_key_without_shrinkage | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 4.361 ms | 61 - 253 MB | NPU
|
| 89 |
+
| encode_key_without_shrinkage | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 3.559 ms | 0 - 226 MB | NPU
|
| 90 |
+
| encode_key_without_shrinkage | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 5.373 ms | 19 - 241 MB | NPU
|
| 91 |
+
| encode_key_without_shrinkage | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 33.113 ms | 9 - 193 MB | NPU
|
| 92 |
+
| encode_key_without_shrinkage | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 7.509 ms | 20 - 23 MB | NPU
|
| 93 |
+
| encode_key_without_shrinkage | TFLITE | float | Qualcomm® SA8775P | 11.353 ms | 20 - 206 MB | NPU
|
| 94 |
+
| encode_key_without_shrinkage | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 14.098 ms | 20 - 222 MB | NPU
|
| 95 |
+
| encode_key_without_shrinkage | TFLITE | float | Qualcomm® SA7255P | 33.113 ms | 9 - 193 MB | NPU
|
| 96 |
+
| encode_key_without_shrinkage | TFLITE | float | Qualcomm® SA8295P | 12.71 ms | 20 - 186 MB | NPU
|
| 97 |
+
| encode_key_without_shrinkage | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 4.413 ms | 18 - 209 MB | NPU
|
| 98 |
+
| encode_value | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 6.596 ms | 0 - 166 MB | NPU
|
| 99 |
+
| encode_value | ONNX | float | Snapdragon® X2 Elite | 6.023 ms | 39 - 39 MB | NPU
|
| 100 |
+
| encode_value | ONNX | float | Snapdragon® X Elite | 12.457 ms | 38 - 38 MB | NPU
|
| 101 |
+
| encode_value | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 9.879 ms | 12 - 220 MB | NPU
|
| 102 |
+
| encode_value | ONNX | float | Qualcomm® QCS8550 (Proxy) | 15.632 ms | 0 - 45 MB | NPU
|
| 103 |
+
| encode_value | ONNX | float | Qualcomm® QCS9075 | 18.755 ms | 6 - 14 MB | NPU
|
| 104 |
+
| encode_value | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 8.398 ms | 7 - 175 MB | NPU
|
| 105 |
+
| encode_value | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 6.27 ms | 1 - 209 MB | NPU
|
| 106 |
+
| encode_value | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 8.318 ms | 3 - 293 MB | NPU
|
| 107 |
+
| encode_value | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 51.385 ms | 3 - 225 MB | NPU
|
| 108 |
+
| encode_value | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 11.877 ms | 3 - 6 MB | NPU
|
| 109 |
+
| encode_value | TFLITE | float | Qualcomm® SA8775P | 17.446 ms | 3 - 220 MB | NPU
|
| 110 |
+
| encode_value | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 22.471 ms | 3 - 286 MB | NPU
|
| 111 |
+
| encode_value | TFLITE | float | Qualcomm® SA7255P | 51.385 ms | 3 - 225 MB | NPU
|
| 112 |
+
| encode_value | TFLITE | float | Qualcomm® SA8295P | 19.759 ms | 3 - 233 MB | NPU
|
| 113 |
+
| encode_value | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 7.187 ms | 2 - 203 MB | NPU
|
| 114 |
+
| segment | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 10.468 ms | 8 - 221 MB | NPU
|
| 115 |
+
| segment | ONNX | float | Snapdragon® X2 Elite | 11.788 ms | 38 - 38 MB | NPU
|
| 116 |
+
| segment | ONNX | float | Snapdragon® X Elite | 23.51 ms | 37 - 37 MB | NPU
|
| 117 |
+
| segment | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 17.508 ms | 0 - 241 MB | NPU
|
| 118 |
+
| segment | ONNX | float | Qualcomm® QCS8550 (Proxy) | 23.222 ms | 0 - 45 MB | NPU
|
| 119 |
+
| segment | ONNX | float | Qualcomm® QCS9075 | 39.325 ms | 21 - 45 MB | NPU
|
| 120 |
+
| segment | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 13.838 ms | 5 - 221 MB | NPU
|
| 121 |
+
| segment | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 10.8 ms | 0 - 227 MB | NPU
|
| 122 |
+
| segment | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 17.715 ms | 0 - 288 MB | NPU
|
| 123 |
+
| segment | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 103.692 ms | 2 - 197 MB | NPU
|
| 124 |
+
| segment | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 23.394 ms | 2 - 4 MB | NPU
|
| 125 |
+
| segment | TFLITE | float | Qualcomm® SA8775P | 33.678 ms | 2 - 202 MB | NPU
|
| 126 |
+
| segment | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 54.382 ms | 2 - 286 MB | NPU
|
| 127 |
+
| segment | TFLITE | float | Qualcomm® SA7255P | 103.692 ms | 2 - 197 MB | NPU
|
| 128 |
+
| segment | TFLITE | float | Qualcomm® SA8295P | 38.758 ms | 2 - 210 MB | NPU
|
| 129 |
+
| segment | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 13.3 ms | 1 - 206 MB | NPU
|
| 130 |
|
| 131 |
## License
|
| 132 |
* The license for the original implementation of Track-Anything can be found
|
release_assets.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"version":"0.50.0","precisions":{"float":{"universal_assets":{"tflite":{"tool_versions":{"qairt":"2.43.0.260127150333_193827","tflite":"2.17.0"},"download_url":"https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/track_anything/releases/v0.50.0/track_anything-tflite-float.zip"},"onnx":{"tool_versions":{"qairt":"2.42.0.251225135753_193295","onnx_runtime":"1.24.1"},"download_url":"https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/track_anything/releases/v0.50.0/track_anything-onnx-float.zip"}}}}}
|