EfficientNet-B0: Optimized for Qualcomm Devices
EfficientNetB0 is a machine learning model that can classify images from the Imagenet dataset. It can also be used as a backbone in building more complex models for specific use cases.
This is based on the implementation of EfficientNet-B0 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 | w8a16_mixed_int16 | 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 | w8a16_mixed_int16 | Universal | QAIRT 2.45 | Download |
| TFLITE | float | Universal | QAIRT 2.45 | Download |
For more device-specific assets and performance metrics, visit EfficientNet-B0 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 EfficientNet-B0 on GitHub for usage instructions.
Model Details
Model Type: Model_use_case.image_classification
Model Stats:
- Input resolution: 224x224
- Model checkpoint: Imagenet
- Model size (float): 20.1 MB
- Model size (w8a16): 6.99 MB
- Number of parameters: 5.27M
Performance Summary
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit |
|---|---|---|---|---|---|---|
| EfficientNet-B0 | ONNX | float | Snapdragon® X2 Elite | 0.656 ms | 2 - 2 MB | NPU |
| EfficientNet-B0 | ONNX | float | Snapdragon® X Elite | 1.315 ms | 14 - 14 MB | NPU |
| EfficientNet-B0 | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 0.904 ms | 0 - 69 MB | NPU |
| EfficientNet-B0 | ONNX | float | Snapdragon® 8 Gen 1 Mobile | 2.88 ms | 1 - 78 MB | NPU |
| EfficientNet-B0 | ONNX | float | Qualcomm® Dragonwing™ IQ-8275 | 1.605 ms | 1 - 4 MB | NPU |
| EfficientNet-B0 | ONNX | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.259 ms | 0 - 16 MB | NPU |
| EfficientNet-B0 | ONNX | float | Qualcomm® QCS8450 | 2.88 ms | 1 - 78 MB | NPU |
| EfficientNet-B0 | ONNX | float | Qualcomm® Dragonwing™ IQ-9075 | 1.628 ms | 1 - 3 MB | NPU |
| EfficientNet-B0 | ONNX | float | Qualcomm® Dragonwing™ IQ-X7181 | 1.315 ms | 14 - 14 MB | NPU |
| EfficientNet-B0 | ONNX | float | Qualcomm® Dragonwing™ Q-8750 | 0.697 ms | 0 - 38 MB | NPU |
| EfficientNet-B0 | ONNX | float | Snapdragon® 8 Elite Mobile | 0.697 ms | 0 - 38 MB | NPU |
| EfficientNet-B0 | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.552 ms | 0 - 43 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16 | Snapdragon® X2 Elite | 0.571 ms | 1 - 1 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16 | Snapdragon® X Elite | 1.486 ms | 6 - 6 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16 | Snapdragon® 8 Gen 3 Mobile | 0.949 ms | 0 - 82 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16 | Snapdragon® 8 Gen 1 Mobile | 1.659 ms | 0 - 83 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16 | Qualcomm® Dragonwing™ QCS6490 | 3.816 ms | 0 - 3 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16 | Qualcomm® Dragonwing™ IQ-8275 | 1.363 ms | 0 - 4 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.425 ms | 0 - 102 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16 | Qualcomm® QCS8450 | 1.659 ms | 0 - 83 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16 | Qualcomm® Dragonwing™ IQ-9075 | 1.576 ms | 0 - 3 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16 | Qualcomm® Dragonwing™ IQ-X7181 | 1.486 ms | 6 - 6 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16 | Qualcomm® Dragonwing™ Q-6690 | 5.691 ms | 0 - 65 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16 | Qualcomm® Dragonwing™ Q-7790 | 1.444 ms | 2 - 60 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16 | Qualcomm® Dragonwing™ Q-8750 | 0.668 ms | 0 - 57 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16 | Snapdragon® 8 Elite Mobile | 0.668 ms | 0 - 57 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | 0.555 ms | 0 - 60 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16 | Snapdragon® 7 Gen 4 Mobile | 1.444 ms | 2 - 60 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16_mixed_int16 | Snapdragon® X2 Elite | 0.584 ms | 1 - 1 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16_mixed_int16 | Snapdragon® X Elite | 1.491 ms | 6 - 6 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16_mixed_int16 | Snapdragon® 8 Gen 3 Mobile | 0.951 ms | 0 - 84 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16_mixed_int16 | Qualcomm® Dragonwing™ IQ-8275 | 1.365 ms | 0 - 4 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16_mixed_int16 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.432 ms | 0 - 9 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16_mixed_int16 | Qualcomm® Dragonwing™ IQ-9075 | 1.574 ms | 0 - 3 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16_mixed_int16 | Qualcomm® Dragonwing™ IQ-X7181 | 1.491 ms | 6 - 6 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16_mixed_int16 | Qualcomm® Dragonwing™ Q-6690 | 5.892 ms | 0 - 65 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16_mixed_int16 | Qualcomm® Dragonwing™ Q-7790 | 1.475 ms | 0 - 59 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16_mixed_int16 | Qualcomm® Dragonwing™ Q-8750 | 0.68 ms | 0 - 63 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16_mixed_int16 | Snapdragon® 8 Elite Mobile | 0.68 ms | 0 - 63 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16_mixed_int16 | Snapdragon® 8 Elite Gen 5 Mobile | 0.553 ms | 0 - 63 MB | NPU |
| EfficientNet-B0 | ONNX | w8a16_mixed_int16 | Snapdragon® 7 Gen 4 Mobile | 1.475 ms | 0 - 59 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Snapdragon® X2 Elite | 0.844 ms | 1 - 1 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Snapdragon® X Elite | 1.753 ms | 1 - 1 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 1.075 ms | 0 - 65 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Snapdragon® 8 Gen 1 Mobile | 3.614 ms | 0 - 74 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-8275 | 1.783 ms | 1 - 4 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.562 ms | 0 - 7 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Qualcomm® SA8775P | 2.009 ms | 1 - 40 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Qualcomm® SA8650P | 2.009 ms | 1 - 40 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Qualcomm® SA8255P | 2.009 ms | 1 - 40 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Qualcomm® QCS8450 | 3.614 ms | 0 - 74 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-9075 | 1.871 ms | 1 - 3 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-X7181 | 1.753 ms | 1 - 1 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Qualcomm® Dragonwing™ Q-8750 | 0.806 ms | 1 - 36 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Qualcomm® SA7255P | 4.947 ms | 1 - 36 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Qualcomm® SA8295P | 3.618 ms | 0 - 43 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Snapdragon® 8 Elite Mobile | 0.806 ms | 1 - 36 MB | NPU |
| EfficientNet-B0 | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.613 ms | 1 - 42 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Snapdragon® X2 Elite | 0.803 ms | 0 - 0 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Snapdragon® X Elite | 1.862 ms | 0 - 0 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Snapdragon® 8 Gen 3 Mobile | 1.132 ms | 0 - 66 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Snapdragon® 8 Gen 1 Mobile | 1.97 ms | 0 - 68 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ QCS6490 | 5.057 ms | 0 - 2 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ IQ-8275 | 1.569 ms | 0 - 3 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.672 ms | 0 - 2 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Qualcomm® SA8775P | 1.947 ms | 0 - 52 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Qualcomm® SA8650P | 1.947 ms | 0 - 52 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Qualcomm® SA8255P | 1.947 ms | 0 - 52 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Qualcomm® QCS8450 | 1.97 ms | 0 - 68 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ IQ-9075 | 1.783 ms | 0 - 2 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ IQ-X7181 | 1.862 ms | 0 - 0 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ Q-6690 | 6.469 ms | 0 - 164 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ Q-7790 | 1.708 ms | 0 - 48 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ Q-8750 | 0.773 ms | 0 - 46 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Qualcomm® SA7255P | 3.281 ms | 0 - 47 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Qualcomm® SA8295P | 2.375 ms | 0 - 46 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Snapdragon® 8 Elite Mobile | 0.773 ms | 0 - 46 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | 0.645 ms | 0 - 47 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16 | Snapdragon® 7 Gen 4 Mobile | 1.708 ms | 0 - 48 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Snapdragon® X2 Elite | 0.816 ms | 0 - 0 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Snapdragon® X Elite | 1.9 ms | 0 - 0 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Snapdragon® 8 Gen 3 Mobile | 1.134 ms | 0 - 66 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Qualcomm® Dragonwing™ IQ-8275 | 1.586 ms | 0 - 3 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.675 ms | 0 - 2 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Qualcomm® SA8775P | 1.951 ms | 0 - 51 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Qualcomm® SA8650P | 1.951 ms | 0 - 51 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Qualcomm® SA8255P | 1.951 ms | 0 - 51 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Qualcomm® Dragonwing™ IQ-9075 | 1.803 ms | 0 - 2 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Qualcomm® Dragonwing™ IQ-X7181 | 1.9 ms | 0 - 0 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Qualcomm® Dragonwing™ Q-6690 | 6.742 ms | 0 - 169 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Qualcomm® Dragonwing™ Q-7790 | 1.718 ms | 0 - 48 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Qualcomm® Dragonwing™ Q-8750 | 0.779 ms | 0 - 47 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Qualcomm® SA7255P | 3.3 ms | 0 - 48 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Snapdragon® 8 Elite Mobile | 0.779 ms | 0 - 47 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Snapdragon® 8 Elite Gen 5 Mobile | 0.634 ms | 0 - 47 MB | NPU |
| EfficientNet-B0 | QNN_DLC | w8a16_mixed_int16 | Snapdragon® 7 Gen 4 Mobile | 1.718 ms | 0 - 48 MB | NPU |
| EfficientNet-B0 | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 1.068 ms | 0 - 65 MB | NPU |
| EfficientNet-B0 | TFLITE | float | Snapdragon® 8 Gen 1 Mobile | 3.59 ms | 0 - 80 MB | NPU |
| EfficientNet-B0 | TFLITE | float | Qualcomm® Dragonwing™ IQ-8275 | 1.795 ms | 0 - 17 MB | NPU |
| EfficientNet-B0 | TFLITE | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.532 ms | 0 - 2 MB | NPU |
| EfficientNet-B0 | TFLITE | float | Qualcomm® SA8775P | 2.025 ms | 0 - 45 MB | NPU |
| EfficientNet-B0 | TFLITE | float | Qualcomm® SA8650P | 2.025 ms | 0 - 45 MB | NPU |
| EfficientNet-B0 | TFLITE | float | Qualcomm® SA8255P | 2.025 ms | 0 - 45 MB | NPU |
| EfficientNet-B0 | TFLITE | float | Qualcomm® QCS8450 | 3.59 ms | 0 - 80 MB | NPU |
| EfficientNet-B0 | TFLITE | float | Qualcomm® Dragonwing™ IQ-9075 | 1.882 ms | 0 - 16 MB | NPU |
| EfficientNet-B0 | TFLITE | float | Qualcomm® Dragonwing™ Q-8750 | 0.816 ms | 0 - 46 MB | NPU |
| EfficientNet-B0 | TFLITE | float | Qualcomm® SA7255P | 4.922 ms | 0 - 41 MB | NPU |
| EfficientNet-B0 | TFLITE | float | Qualcomm® SA8295P | 3.634 ms | 0 - 48 MB | NPU |
| EfficientNet-B0 | TFLITE | float | Snapdragon® 8 Elite Mobile | 0.816 ms | 0 - 46 MB | NPU |
| EfficientNet-B0 | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.611 ms | 0 - 41 MB | NPU |
License
- The license for the original implementation of EfficientNet-B0 can be found here.
References
- EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks
- Source Model Implementation
Community
- Join our AI Hub Slack community to collaborate, post questions and learn more about on-device AI.
- For questions or feedback please reach out to us.
