Video Classification
PyTorch
backbone
android

ResNet-3D: Optimized for Qualcomm Devices

ResNet 3D is a network with 3D convolutions used for video understanding.

This is based on the implementation of ResNet-3D 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 w8a8 Universal QAIRT 2.45, ONNX Runtime 1.27.1 Download
QNN_DLC float 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-3D 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-3D 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): 127 MB
  • Model size (w8a8): 32.1 MB
  • Number of parameters: 33.4M

Performance Summary

Model Runtime Precision Chipset Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit
ResNet-3D ONNX float Snapdragon® X2 Elite 7.079 ms 2 - 2 MB NPU
ResNet-3D ONNX float Snapdragon® X Elite 13.45 ms 64 - 64 MB NPU
ResNet-3D ONNX float Snapdragon® 8 Gen 3 Mobile 9.592 ms 0 - 243 MB NPU
ResNet-3D ONNX float Snapdragon® 8 Gen 1 Mobile 25.797 ms 1 - 225 MB NPU
ResNet-3D ONNX float Qualcomm® Dragonwing™ IQ-8275 27.183 ms 2 - 8 MB NPU
ResNet-3D ONNX float Qualcomm® Dragonwing™ QCS8550 (Proxy) 13.15 ms 0 - 80 MB NPU
ResNet-3D ONNX float Qualcomm® QCS8450 25.797 ms 1 - 225 MB NPU
ResNet-3D ONNX float Qualcomm® Dragonwing™ IQ-9075 24.691 ms 2 - 7 MB NPU
ResNet-3D ONNX float Qualcomm® Dragonwing™ IQ-X7181 13.45 ms 64 - 64 MB NPU
ResNet-3D ONNX float Qualcomm® Dragonwing™ Q-8750 7.887 ms 1 - 189 MB NPU
ResNet-3D ONNX float Snapdragon® 8 Elite Mobile 7.887 ms 1 - 189 MB NPU
ResNet-3D ONNX float Snapdragon® 8 Elite Gen 5 Mobile 6.152 ms 0 - 195 MB NPU
ResNet-3D ONNX w8a8 Snapdragon® X2 Elite 3.291 ms 2 - 2 MB NPU
ResNet-3D ONNX w8a8 Snapdragon® X Elite 4.577 ms 33 - 33 MB NPU
ResNet-3D ONNX w8a8 Snapdragon® 8 Gen 3 Mobile 3.338 ms 0 - 240 MB NPU
ResNet-3D ONNX w8a8 Snapdragon® 8 Gen 1 Mobile 6.52 ms 1 - 244 MB NPU
ResNet-3D ONNX w8a8 Qualcomm® Dragonwing™ QCS6490 26.01 ms 0 - 3 MB NPU
ResNet-3D ONNX w8a8 Qualcomm® Dragonwing™ IQ-8275 4.986 ms 0 - 4 MB NPU
ResNet-3D ONNX w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 4.494 ms 0 - 42 MB NPU
ResNet-3D ONNX w8a8 Qualcomm® QCS8450 6.52 ms 1 - 244 MB NPU
ResNet-3D ONNX w8a8 Qualcomm® Dragonwing™ IQ-9075 4.44 ms 0 - 3 MB NPU
ResNet-3D ONNX w8a8 Qualcomm® Dragonwing™ IQ-X7181 4.577 ms 33 - 33 MB NPU
ResNet-3D ONNX w8a8 Qualcomm® Dragonwing™ Q-6690 89.831 ms 1 - 178 MB NPU
ResNet-3D ONNX w8a8 Qualcomm® Dragonwing™ Q-7790 7.787 ms 0 - 168 MB NPU
ResNet-3D ONNX w8a8 Qualcomm® Dragonwing™ Q-8750 2.737 ms 0 - 180 MB NPU
ResNet-3D ONNX w8a8 Snapdragon® 8 Elite Mobile 2.737 ms 0 - 180 MB NPU
ResNet-3D ONNX w8a8 Snapdragon® 8 Elite Gen 5 Mobile 1.998 ms 0 - 180 MB NPU
ResNet-3D ONNX w8a8 Snapdragon® 7 Gen 4 Mobile 7.787 ms 0 - 168 MB NPU
ResNet-3D QNN_DLC float Snapdragon® X2 Elite 7.424 ms 2 - 2 MB NPU
ResNet-3D QNN_DLC float Snapdragon® X Elite 14.16 ms 2 - 2 MB NPU
ResNet-3D QNN_DLC float Snapdragon® 8 Gen 3 Mobile 9.723 ms 0 - 274 MB NPU
ResNet-3D QNN_DLC float Snapdragon® 8 Gen 1 Mobile 28.042 ms 2 - 240 MB NPU
ResNet-3D QNN_DLC float Qualcomm® Dragonwing™ IQ-8275 27.084 ms 2 - 7 MB NPU
ResNet-3D QNN_DLC float Qualcomm® Dragonwing™ QCS8550 (Proxy) 13.347 ms 2 - 9 MB NPU
ResNet-3D QNN_DLC float Qualcomm® SA8775P 24.035 ms 1 - 199 MB NPU
ResNet-3D QNN_DLC float Qualcomm® SA8650P 24.035 ms 1 - 199 MB NPU
ResNet-3D QNN_DLC float Qualcomm® SA8255P 24.035 ms 1 - 199 MB NPU
ResNet-3D QNN_DLC float Qualcomm® QCS8450 28.042 ms 2 - 240 MB NPU
ResNet-3D QNN_DLC float Qualcomm® Dragonwing™ IQ-9075 25.007 ms 2 - 6 MB NPU
ResNet-3D QNN_DLC float Qualcomm® Dragonwing™ IQ-X7181 14.16 ms 2 - 2 MB NPU
ResNet-3D QNN_DLC float Qualcomm® Dragonwing™ Q-8750 7.489 ms 2 - 204 MB NPU
ResNet-3D QNN_DLC float Qualcomm® SA7255P 91.773 ms 1 - 213 MB NPU
ResNet-3D QNN_DLC float Qualcomm® SA8295P 26.42 ms 1 - 194 MB NPU
ResNet-3D QNN_DLC float Snapdragon® 8 Elite Mobile 7.489 ms 2 - 204 MB NPU
ResNet-3D QNN_DLC float Snapdragon® 8 Elite Gen 5 Mobile 5.928 ms 2 - 216 MB NPU
ResNet-3D QNN_DLC w8a8 Snapdragon® X2 Elite 2.445 ms 1 - 1 MB NPU
ResNet-3D QNN_DLC w8a8 Snapdragon® X Elite 4.906 ms 1 - 1 MB NPU
ResNet-3D QNN_DLC w8a8 Snapdragon® 8 Gen 3 Mobile 3.376 ms 0 - 125 MB NPU
ResNet-3D QNN_DLC w8a8 Snapdragon® 8 Gen 1 Mobile 6.06 ms 1 - 125 MB NPU
ResNet-3D QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS6490 26.436 ms 3 - 5 MB NPU
ResNet-3D QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-8275 4.997 ms 1 - 4 MB NPU
ResNet-3D QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 4.529 ms 1 - 114 MB NPU
ResNet-3D QNN_DLC w8a8 Qualcomm® SA8775P 4.612 ms 1 - 62 MB NPU
ResNet-3D QNN_DLC w8a8 Qualcomm® SA8650P 4.612 ms 1 - 62 MB NPU
ResNet-3D QNN_DLC w8a8 Qualcomm® SA8255P 4.612 ms 1 - 62 MB NPU
ResNet-3D QNN_DLC w8a8 Qualcomm® QCS8450 6.06 ms 1 - 125 MB NPU
ResNet-3D QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-9075 4.415 ms 3 - 5 MB NPU
ResNet-3D QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-X7181 4.906 ms 1 - 1 MB NPU
ResNet-3D QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-6690 92.47 ms 1 - 186 MB NPU
ResNet-3D QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-7790 7.753 ms 1 - 173 MB NPU
ResNet-3D QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-8750 2.677 ms 0 - 182 MB NPU
ResNet-3D QNN_DLC w8a8 Qualcomm® SA7255P 14.996 ms 1 - 61 MB NPU
ResNet-3D QNN_DLC w8a8 Qualcomm® SA8295P 8.306 ms 1 - 58 MB NPU
ResNet-3D QNN_DLC w8a8 Snapdragon® 8 Elite Mobile 2.677 ms 0 - 182 MB NPU
ResNet-3D QNN_DLC w8a8 Snapdragon® 8 Elite Gen 5 Mobile 1.939 ms 1 - 65 MB NPU
ResNet-3D QNN_DLC w8a8 Snapdragon® 7 Gen 4 Mobile 7.753 ms 1 - 173 MB NPU
ResNet-3D TFLITE float Snapdragon® 8 Gen 3 Mobile 211.824 ms 0 - 275 MB NPU
ResNet-3D TFLITE float Snapdragon® 8 Gen 1 Mobile 308.04 ms 0 - 261 MB NPU
ResNet-3D TFLITE float Qualcomm® Dragonwing™ IQ-8275 271.257 ms 0 - 70 MB NPU
ResNet-3D TFLITE float Qualcomm® Dragonwing™ QCS8550 (Proxy) 270.601 ms 0 - 2 MB NPU
ResNet-3D TFLITE float Qualcomm® SA8775P 287.875 ms 0 - 213 MB NPU
ResNet-3D TFLITE float Qualcomm® SA8650P 287.875 ms 0 - 213 MB NPU
ResNet-3D TFLITE float Qualcomm® SA8255P 287.875 ms 0 - 213 MB NPU
ResNet-3D TFLITE float Qualcomm® QCS8450 308.04 ms 0 - 261 MB NPU
ResNet-3D TFLITE float Qualcomm® Dragonwing™ IQ-9075 295.6 ms 0 - 70 MB NPU
ResNet-3D TFLITE float Qualcomm® Dragonwing™ Q-8750 188.745 ms 0 - 225 MB NPU
ResNet-3D TFLITE float Qualcomm® SA7255P 563.821 ms 1 - 220 MB NPU
ResNet-3D TFLITE float Qualcomm® SA8295P 326.664 ms 0 - 214 MB NPU
ResNet-3D TFLITE float Snapdragon® 8 Elite Mobile 188.745 ms 0 - 225 MB NPU
ResNet-3D TFLITE float Snapdragon® 8 Elite Gen 5 Mobile 175.587 ms 0 - 227 MB NPU
ResNet-3D TFLITE w8a8 Snapdragon® 8 Gen 3 Mobile 179.116 ms 0 - 313 MB NPU
ResNet-3D TFLITE w8a8 Snapdragon® 8 Gen 1 Mobile 252.719 ms 0 - 311 MB NPU
ResNet-3D TFLITE w8a8 Qualcomm® Dragonwing™ QCS6490 1736.117 ms 693 - 1058 MB NPU
ResNet-3D TFLITE w8a8 Qualcomm® Dragonwing™ IQ-8275 238.914 ms 0 - 69 MB NPU
ResNet-3D TFLITE w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 234.346 ms 0 - 3 MB NPU
ResNet-3D TFLITE w8a8 Qualcomm® SA8775P 247.045 ms 1 - 299 MB NPU
ResNet-3D TFLITE w8a8 Qualcomm® SA8650P 247.045 ms 1 - 299 MB NPU
ResNet-3D TFLITE w8a8 Qualcomm® SA8255P 247.045 ms 1 - 299 MB NPU
ResNet-3D TFLITE w8a8 Qualcomm® QCS8450 252.719 ms 0 - 311 MB NPU
ResNet-3D TFLITE w8a8 Qualcomm® Dragonwing™ IQ-9075 249.307 ms 0 - 68 MB NPU
ResNet-3D TFLITE w8a8 Qualcomm® Dragonwing™ Q-6690 1480.63 ms 660 - 813 MB NPU
ResNet-3D TFLITE w8a8 Qualcomm® Dragonwing™ Q-7790 1175.734 ms 654 - 705 MB NPU
ResNet-3D TFLITE w8a8 Qualcomm® Dragonwing™ Q-8750 131.864 ms 0 - 384 MB NPU
ResNet-3D TFLITE w8a8 Qualcomm® SA7255P 498.163 ms 0 - 253 MB NPU
ResNet-3D TFLITE w8a8 Qualcomm® SA8295P 261.042 ms 0 - 252 MB NPU
ResNet-3D TFLITE w8a8 Snapdragon® 8 Elite Mobile 131.864 ms 0 - 384 MB NPU
ResNet-3D TFLITE w8a8 Snapdragon® 8 Elite Gen 5 Mobile 163.86 ms 0 - 313 MB NPU
ResNet-3D TFLITE w8a8 Snapdragon® 7 Gen 4 Mobile 1175.734 ms 654 - 705 MB NPU

License

  • The license for the original implementation of ResNet-3D can be found here.

References

Community

Downloads last month

-

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

Paper for qualcomm/ResNet-3D