Hugging Face's logo

qualcomm
/
ResNet-Mixed-Convolution

Video Classification
PyTorch
backbone
android
Model card Files
xet
 Community
1

ResNet-Mixed-Convolution: Optimized for Qualcomm Devices

ResNet Mixed Convolutions is a network with a mixture of 2D and 3D convolutions used for video understanding.

This is based on the implementation of ResNet-Mixed-Convolution found here. This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the Qualcomm® AI Hub Models library to export with custom configurations. More details on model performance across various devices, can be found here.

Qualcomm AI Hub Models uses Qualcomm AI Hub Workbench to compile, profile, and evaluate this model. Sign up 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
ONNX float Universal QAIRT 2.45, ONNX Runtime 1.27.1 Download
ONNX w8a16 Universal QAIRT 2.45, ONNX Runtime 1.27.1 Download
ONNX w8a8 Universal QAIRT 2.45, ONNX Runtime 1.27.1 Download
QNN_DLC float Universal QAIRT 2.45 Download
QNN_DLC w8a16 Universal QAIRT 2.45 Download
QNN_DLC w8a8 Universal QAIRT 2.45 Download
TFLITE float Universal QAIRT 2.45 Download
TFLITE w8a8 Universal QAIRT 2.45 Download

For more device-specific assets and performance metrics, visit ResNet-Mixed-Convolution on Qualcomm® AI Hub.

Option 2: Export with Custom Configurations

Use the Qualcomm® AI Hub Models 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 ResNet-Mixed-Convolution on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.video_classification

Model Stats:

  • Input resolution: 112x112
  • Model checkpoint: Kinetics-400
  • Model size (float): 44.6 MB
  • Model size (w8a16): 11.5 MB
  • Number of parameters: 11.7M

Performance Summary

Model Runtime Precision Chipset Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit
ResNet-Mixed-Convolution ONNX float Snapdragon® X2 Elite 7.2 ms 2 - 2 MB NPU
ResNet-Mixed-Convolution ONNX float Snapdragon® X Elite 13.61 ms 22 - 22 MB NPU
ResNet-Mixed-Convolution ONNX float Snapdragon® 8 Gen 3 Mobile 9.64 ms 2 - 262 MB NPU
ResNet-Mixed-Convolution ONNX float Snapdragon® 8 Gen 1 Mobile 25.919 ms 2 - 225 MB NPU
ResNet-Mixed-Convolution ONNX float Qualcomm® Dragonwing™ IQ-8275 28.462 ms 2 - 6 MB NPU
ResNet-Mixed-Convolution ONNX float Qualcomm® Dragonwing™ QCS8550 (Proxy) 13.055 ms 0 - 183 MB NPU
ResNet-Mixed-Convolution ONNX float Qualcomm® QCS8450 25.919 ms 2 - 225 MB NPU
ResNet-Mixed-Convolution ONNX float Qualcomm® Dragonwing™ IQ-9075 25.621 ms 2 - 5 MB NPU
ResNet-Mixed-Convolution ONNX float Qualcomm® Dragonwing™ IQ-X7181 13.61 ms 22 - 22 MB NPU
ResNet-Mixed-Convolution ONNX float Qualcomm® Dragonwing™ Q-8750 7.866 ms 1 - 211 MB NPU
ResNet-Mixed-Convolution ONNX float Snapdragon® 8 Elite Mobile 7.866 ms 1 - 211 MB NPU
ResNet-Mixed-Convolution ONNX float Snapdragon® 8 Elite Gen 5 Mobile 5.818 ms 2 - 217 MB NPU
ResNet-Mixed-Convolution ONNX w8a16 Snapdragon® X2 Elite 4.673 ms 1 - 1 MB NPU
ResNet-Mixed-Convolution ONNX w8a16 Snapdragon® X Elite 8.971 ms 11 - 11 MB NPU
ResNet-Mixed-Convolution ONNX w8a16 Snapdragon® 8 Gen 3 Mobile 6.435 ms 1 - 244 MB NPU
ResNet-Mixed-Convolution ONNX w8a16 Snapdragon® 8 Gen 1 Mobile 12.503 ms 1 - 243 MB NPU
ResNet-Mixed-Convolution ONNX w8a16 Qualcomm® Dragonwing™ QCS6490 50.597 ms 1 - 4 MB NPU
ResNet-Mixed-Convolution ONNX w8a16 Qualcomm® Dragonwing™ IQ-8275 9.553 ms 1 - 5 MB NPU
ResNet-Mixed-Convolution ONNX w8a16 Qualcomm® Dragonwing™ QCS8550 (Proxy) 8.605 ms 0 - 15 MB NPU
ResNet-Mixed-Convolution ONNX w8a16 Qualcomm® QCS8450 12.503 ms 1 - 243 MB NPU
ResNet-Mixed-Convolution ONNX w8a16 Qualcomm® Dragonwing™ IQ-9075 8.344 ms 1 - 4 MB NPU
ResNet-Mixed-Convolution ONNX w8a16 Qualcomm® Dragonwing™ IQ-X7181 8.971 ms 11 - 11 MB NPU
ResNet-Mixed-Convolution ONNX w8a16 Qualcomm® Dragonwing™ Q-6690 177.681 ms 1 - 201 MB NPU
ResNet-Mixed-Convolution ONNX w8a16 Qualcomm® Dragonwing™ Q-7790 15.248 ms 1 - 197 MB NPU
ResNet-Mixed-Convolution ONNX w8a16 Qualcomm® Dragonwing™ Q-8750 5.219 ms 1 - 185 MB NPU
ResNet-Mixed-Convolution ONNX w8a16 Snapdragon® 8 Elite Mobile 5.219 ms 1 - 185 MB NPU
ResNet-Mixed-Convolution ONNX w8a16 Snapdragon® 8 Elite Gen 5 Mobile 3.662 ms 1 - 194 MB NPU
ResNet-Mixed-Convolution ONNX w8a16 Snapdragon® 7 Gen 4 Mobile 15.248 ms 1 - 197 MB NPU
ResNet-Mixed-Convolution ONNX w8a8 Snapdragon® X2 Elite 2.333 ms 2 - 2 MB NPU
ResNet-Mixed-Convolution ONNX w8a8 Snapdragon® X Elite 4.917 ms 12 - 12 MB NPU
ResNet-Mixed-Convolution ONNX w8a8 Snapdragon® 8 Gen 3 Mobile 3.396 ms 0 - 219 MB NPU
ResNet-Mixed-Convolution ONNX w8a8 Snapdragon® 8 Gen 1 Mobile 6.997 ms 1 - 217 MB NPU
ResNet-Mixed-Convolution ONNX w8a8 Qualcomm® Dragonwing™ QCS6490 24.708 ms 1 - 3 MB NPU
ResNet-Mixed-Convolution ONNX w8a8 Qualcomm® Dragonwing™ IQ-8275 5.175 ms 1 - 4 MB NPU
ResNet-Mixed-Convolution ONNX w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 4.737 ms 0 - 175 MB NPU
ResNet-Mixed-Convolution ONNX w8a8 Qualcomm® QCS8450 6.997 ms 1 - 217 MB NPU
ResNet-Mixed-Convolution ONNX w8a8 Qualcomm® Dragonwing™ IQ-9075 4.553 ms 1 - 3 MB NPU
ResNet-Mixed-Convolution ONNX w8a8 Qualcomm® Dragonwing™ IQ-X7181 4.917 ms 12 - 12 MB NPU
ResNet-Mixed-Convolution ONNX w8a8 Qualcomm® Dragonwing™ Q-6690 95.515 ms 1 - 190 MB NPU
ResNet-Mixed-Convolution ONNX w8a8 Qualcomm® Dragonwing™ Q-7790 7.996 ms 1 - 177 MB NPU
ResNet-Mixed-Convolution ONNX w8a8 Qualcomm® Dragonwing™ Q-8750 2.844 ms 0 - 184 MB NPU
ResNet-Mixed-Convolution ONNX w8a8 Snapdragon® 8 Elite Mobile 2.844 ms 0 - 184 MB NPU
ResNet-Mixed-Convolution ONNX w8a8 Snapdragon® 8 Elite Gen 5 Mobile 1.899 ms 0 - 179 MB NPU
ResNet-Mixed-Convolution ONNX w8a8 Snapdragon® 7 Gen 4 Mobile 7.996 ms 1 - 177 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Snapdragon® X2 Elite 7.743 ms 2 - 2 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Snapdragon® X Elite 14.165 ms 2 - 2 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Snapdragon® 8 Gen 3 Mobile 9.548 ms 2 - 280 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Snapdragon® 8 Gen 1 Mobile 27.563 ms 2 - 240 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Qualcomm® Dragonwing™ IQ-8275 28.052 ms 2 - 7 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Qualcomm® Dragonwing™ QCS8550 (Proxy) 13.327 ms 2 - 5 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Qualcomm® SA8775P 25.19 ms 1 - 224 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Qualcomm® SA8650P 25.19 ms 1 - 224 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Qualcomm® SA8255P 25.19 ms 1 - 224 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Qualcomm® QCS8450 27.563 ms 2 - 240 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Qualcomm® Dragonwing™ IQ-9075 25.859 ms 2 - 6 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Qualcomm® Dragonwing™ IQ-X7181 14.165 ms 2 - 2 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Qualcomm® Dragonwing™ Q-8750 7.634 ms 2 - 228 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Qualcomm® SA7255P 97.14 ms 2 - 223 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Qualcomm® SA8295P 26.84 ms 1 - 196 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Snapdragon® 8 Elite Mobile 7.634 ms 2 - 228 MB NPU
ResNet-Mixed-Convolution QNN_DLC float Snapdragon® 8 Elite Gen 5 Mobile 5.903 ms 2 - 234 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Snapdragon® X2 Elite 5.492 ms 1 - 1 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Snapdragon® X Elite 10.19 ms 1 - 1 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Snapdragon® 8 Gen 3 Mobile 6.776 ms 1 - 243 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Snapdragon® 8 Gen 1 Mobile 13.243 ms 1 - 244 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Qualcomm® Dragonwing™ QCS6490 51.728 ms 1 - 4 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Qualcomm® Dragonwing™ IQ-8275 9.931 ms 1 - 5 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Qualcomm® Dragonwing™ QCS8550 (Proxy) 9.417 ms 1 - 31 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Qualcomm® SA8775P 9.375 ms 1 - 191 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Qualcomm® SA8650P 9.375 ms 1 - 191 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Qualcomm® SA8255P 9.375 ms 1 - 191 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Qualcomm® QCS8450 13.243 ms 1 - 244 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Qualcomm® Dragonwing™ IQ-9075 9.078 ms 3 - 6 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Qualcomm® Dragonwing™ IQ-X7181 10.19 ms 1 - 1 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Qualcomm® Dragonwing™ Q-6690 162.199 ms 1 - 209 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Qualcomm® Dragonwing™ Q-7790 15.84 ms 1 - 198 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Qualcomm® Dragonwing™ Q-8750 5.375 ms 1 - 187 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Qualcomm® SA7255P 30.082 ms 1 - 189 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Qualcomm® SA8295P 16.067 ms 1 - 194 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Snapdragon® 8 Elite Mobile 5.375 ms 1 - 187 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Snapdragon® 8 Elite Gen 5 Mobile 4.121 ms 1 - 195 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a16 Snapdragon® 7 Gen 4 Mobile 15.84 ms 1 - 198 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Snapdragon® X2 Elite 2.585 ms 1 - 1 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Snapdragon® X Elite 5.721 ms 1 - 1 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Snapdragon® 8 Gen 3 Mobile 3.746 ms 0 - 213 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Snapdragon® 8 Gen 1 Mobile 6.82 ms 1 - 216 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS6490 27.217 ms 3 - 5 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-8275 5.605 ms 1 - 4 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 5.203 ms 1 - 22 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Qualcomm® SA8775P 5.327 ms 1 - 175 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Qualcomm® SA8650P 5.327 ms 1 - 175 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Qualcomm® SA8255P 5.327 ms 1 - 175 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Qualcomm® QCS8450 6.82 ms 1 - 216 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-9075 5.103 ms 0 - 3 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-X7181 5.721 ms 1 - 1 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-6690 99.461 ms 1 - 201 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-7790 7.894 ms 1 - 181 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-8750 2.997 ms 0 - 176 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Qualcomm® SA7255P 17.529 ms 1 - 174 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Qualcomm® SA8295P 9.392 ms 1 - 177 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Snapdragon® 8 Elite Mobile 2.997 ms 0 - 176 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Snapdragon® 8 Elite Gen 5 Mobile 1.907 ms 1 - 192 MB NPU
ResNet-Mixed-Convolution QNN_DLC w8a8 Snapdragon® 7 Gen 4 Mobile 7.894 ms 1 - 181 MB NPU
ResNet-Mixed-Convolution TFLITE float Snapdragon® 8 Gen 3 Mobile 220.796 ms 0 - 304 MB NPU
ResNet-Mixed-Convolution TFLITE float Snapdragon® 8 Gen 1 Mobile 337.377 ms 1 - 270 MB NPU
ResNet-Mixed-Convolution TFLITE float Qualcomm® Dragonwing™ IQ-8275 290.055 ms 0 - 29 MB NPU
ResNet-Mixed-Convolution TFLITE float Qualcomm® Dragonwing™ QCS8550 (Proxy) 300.222 ms 0 - 2 MB NPU
ResNet-Mixed-Convolution TFLITE float Qualcomm® SA8775P 308.468 ms 0 - 254 MB NPU
ResNet-Mixed-Convolution TFLITE float Qualcomm® SA8650P 308.468 ms 0 - 254 MB NPU
ResNet-Mixed-Convolution TFLITE float Qualcomm® SA8255P 308.468 ms 0 - 254 MB NPU
ResNet-Mixed-Convolution TFLITE float Qualcomm® QCS8450 337.377 ms 1 - 270 MB NPU
ResNet-Mixed-Convolution TFLITE float Qualcomm® Dragonwing™ IQ-9075 312.716 ms 0 - 28 MB NPU
ResNet-Mixed-Convolution TFLITE float Qualcomm® Dragonwing™ Q-8750 205.897 ms 0 - 255 MB NPU
ResNet-Mixed-Convolution TFLITE float Qualcomm® SA7255P 589.362 ms 0 - 253 MB NPU
ResNet-Mixed-Convolution TFLITE float Qualcomm® SA8295P 355.857 ms 0 - 234 MB NPU
ResNet-Mixed-Convolution TFLITE float Snapdragon® 8 Elite Mobile 205.897 ms 0 - 255 MB NPU
ResNet-Mixed-Convolution TFLITE float Snapdragon® 8 Elite Gen 5 Mobile 196.479 ms 0 - 251 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Snapdragon® 8 Gen 3 Mobile 440.203 ms 0 - 351 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Snapdragon® 8 Gen 1 Mobile 660.525 ms 1 - 347 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Qualcomm® Dragonwing™ QCS6490 1778.952 ms 695 - 1060 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Qualcomm® Dragonwing™ IQ-8275 587.121 ms 0 - 28 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 570.377 ms 0 - 3 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Qualcomm® SA8775P 601.874 ms 1 - 296 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Qualcomm® SA8650P 601.874 ms 1 - 296 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Qualcomm® SA8255P 601.874 ms 1 - 296 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Qualcomm® QCS8450 660.525 ms 1 - 347 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Qualcomm® Dragonwing™ IQ-9075 600.581 ms 0 - 27 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Qualcomm® Dragonwing™ Q-6690 1542.659 ms 653 - 809 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Qualcomm® Dragonwing™ Q-7790 1236.014 ms 638 - 809 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Qualcomm® Dragonwing™ Q-8750 391.517 ms 0 - 279 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Qualcomm® SA7255P 1176.404 ms 0 - 297 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Qualcomm® SA8295P 656.592 ms 0 - 297 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Snapdragon® 8 Elite Mobile 391.517 ms 0 - 279 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Snapdragon® 8 Elite Gen 5 Mobile 433.967 ms 0 - 343 MB NPU
ResNet-Mixed-Convolution TFLITE w8a8 Snapdragon® 7 Gen 4 Mobile 1236.014 ms 638 - 809 MB NPU

License

  • The license for the original implementation of ResNet-Mixed-Convolution can be found here.

References

Community

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
Video Classification
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Paper for qualcomm/ResNet-Mixed-Convolution