---
library_name: pytorch
license: other
tags:
- android
pipeline_tag: image-to-image

---

![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/sesr_m5/web-assets/model_demo.png)

# SESR-M5: Optimized for Qualcomm Devices

SESR M5 performs efficient on-device upscaling of images.

This is based on the implementation of SESR-M5 found [here](https://github.com/quic/aimet-model-zoo/tree/develop/aimet_zoo_torch/sesr).
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/src/qai_hub_models/models/sesr_m5) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).

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.

## 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 2.42, ONNX Runtime 1.24.3 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/sesr_m5/releases/v0.51.0/sesr_m5-onnx-float.zip)
| ONNX | w8a8 | Universal | QAIRT 2.42, ONNX Runtime 1.24.3 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/sesr_m5/releases/v0.51.0/sesr_m5-onnx-w8a8.zip)
| QNN_DLC | float | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/sesr_m5/releases/v0.51.0/sesr_m5-qnn_dlc-float.zip)
| QNN_DLC | w8a8 | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/sesr_m5/releases/v0.51.0/sesr_m5-qnn_dlc-w8a8.zip)
| TFLITE | float | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/sesr_m5/releases/v0.51.0/sesr_m5-tflite-float.zip)
| TFLITE | w8a8 | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/sesr_m5/releases/v0.51.0/sesr_m5-tflite-w8a8.zip)

For more device-specific assets and performance metrics, visit **[SESR-M5 on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/sesr_m5)**.


### Option 2: Export with Custom Configurations

Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/src/qai_hub_models/models/sesr_m5) 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 [SESR-M5 on GitHub](https://github.com/qualcomm/ai-hub-models/blob/main/src/qai_hub_models/models/sesr_m5) for usage instructions.

## Model Details

**Model Type:** Model_use_case.super_resolution

**Model Stats:**
- Model checkpoint: sesr_m5_3x_checkpoint
- Input resolution: 128x128
- Number of parameters: 343K
- Model size (float): 1.32 MB
- Model size (w8a8): 395 KB

## Performance Summary
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|---|---|---|---|---|---|---
| SESR-M5 | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.97 ms | 0 - 25 MB | NPU
| SESR-M5 | ONNX | float | Snapdragon® X2 Elite | 1.146 ms | 6 - 6 MB | NPU
| SESR-M5 | ONNX | float | Snapdragon® X Elite | 2.204 ms | 8 - 8 MB | NPU
| SESR-M5 | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 1.457 ms | 0 - 35 MB | NPU
| SESR-M5 | ONNX | float | Qualcomm® QCS8550 (Proxy) | 2.003 ms | 0 - 6 MB | NPU
| SESR-M5 | ONNX | float | Qualcomm® QCS9075 | 3.204 ms | 6 - 9 MB | NPU
| SESR-M5 | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 1.214 ms | 0 - 21 MB | NPU
| SESR-M5 | ONNX | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.375 ms | 0 - 26 MB | NPU
| SESR-M5 | ONNX | w8a8 | Snapdragon® X2 Elite | 0.407 ms | 0 - 0 MB | NPU
| SESR-M5 | ONNX | w8a8 | Snapdragon® X Elite | 0.915 ms | 3 - 3 MB | NPU
| SESR-M5 | ONNX | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.543 ms | 0 - 33 MB | NPU
| SESR-M5 | ONNX | w8a8 | Qualcomm® QCS6490 | 42.215 ms | 38 - 41 MB | CPU
| SESR-M5 | ONNX | w8a8 | Qualcomm® QCS8550 (Proxy) | 0.788 ms | 0 - 2 MB | NPU
| SESR-M5 | ONNX | w8a8 | Qualcomm® QCS9075 | 0.985 ms | 0 - 3 MB | NPU
| SESR-M5 | ONNX | w8a8 | Qualcomm® QCM6690 | 31.246 ms | 38 - 44 MB | CPU
| SESR-M5 | ONNX | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 0.431 ms | 0 - 27 MB | NPU
| SESR-M5 | ONNX | w8a8 | Snapdragon® 7 Gen 4 Mobile | 23.807 ms | 38 - 45 MB | CPU
| SESR-M5 | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.87 ms | 0 - 27 MB | NPU
| SESR-M5 | QNN_DLC | float | Snapdragon® X2 Elite | 1.202 ms | 0 - 0 MB | NPU
| SESR-M5 | QNN_DLC | float | Snapdragon® X Elite | 2.079 ms | 0 - 0 MB | NPU
| SESR-M5 | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 1.273 ms | 0 - 34 MB | NPU
| SESR-M5 | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 10.526 ms | 0 - 23 MB | NPU
| SESR-M5 | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 1.874 ms | 0 - 5 MB | NPU
| SESR-M5 | QNN_DLC | float | Qualcomm® SA8775P | 3.03 ms | 0 - 25 MB | NPU
| SESR-M5 | QNN_DLC | float | Qualcomm® QCS9075 | 3.054 ms | 0 - 5 MB | NPU
| SESR-M5 | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 2.771 ms | 0 - 37 MB | NPU
| SESR-M5 | QNN_DLC | float | Qualcomm® SA7255P | 10.526 ms | 0 - 23 MB | NPU
| SESR-M5 | QNN_DLC | float | Qualcomm® SA8295P | 3.512 ms | 0 - 22 MB | NPU
| SESR-M5 | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 1.03 ms | 0 - 24 MB | NPU
| SESR-M5 | QNN_DLC | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.296 ms | 0 - 26 MB | NPU
| SESR-M5 | QNN_DLC | w8a8 | Snapdragon® X2 Elite | 0.443 ms | 0 - 0 MB | NPU
| SESR-M5 | QNN_DLC | w8a8 | Snapdragon® X Elite | 0.749 ms | 0 - 0 MB | NPU
| SESR-M5 | QNN_DLC | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.44 ms | 0 - 32 MB | NPU
| SESR-M5 | QNN_DLC | w8a8 | Qualcomm® QCS6490 | 2.675 ms | 0 - 2 MB | NPU
| SESR-M5 | QNN_DLC | w8a8 | Qualcomm® QCS8275 (Proxy) | 1.962 ms | 0 - 23 MB | NPU
| SESR-M5 | QNN_DLC | w8a8 | Qualcomm® QCS8550 (Proxy) | 0.603 ms | 0 - 16 MB | NPU
| SESR-M5 | QNN_DLC | w8a8 | Qualcomm® SA8775P | 0.835 ms | 0 - 26 MB | NPU
| SESR-M5 | QNN_DLC | w8a8 | Qualcomm® QCS9075 | 0.893 ms | 0 - 2 MB | NPU
| SESR-M5 | QNN_DLC | w8a8 | Qualcomm® QCM6690 | 6.815 ms | 0 - 139 MB | NPU
| SESR-M5 | QNN_DLC | w8a8 | Qualcomm® QCS8450 (Proxy) | 0.859 ms | 0 - 34 MB | NPU
| SESR-M5 | QNN_DLC | w8a8 | Qualcomm® SA7255P | 1.962 ms | 0 - 23 MB | NPU
| SESR-M5 | QNN_DLC | w8a8 | Qualcomm® SA8295P | 1.359 ms | 0 - 21 MB | NPU
| SESR-M5 | QNN_DLC | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 0.379 ms | 0 - 23 MB | NPU
| SESR-M5 | QNN_DLC | w8a8 | Snapdragon® 7 Gen 4 Mobile | 0.781 ms | 0 - 22 MB | NPU
| SESR-M5 | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.871 ms | 0 - 28 MB | NPU
| SESR-M5 | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 1.273 ms | 0 - 35 MB | NPU
| SESR-M5 | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 10.579 ms | 3 - 26 MB | NPU
| SESR-M5 | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 1.886 ms | 0 - 5 MB | NPU
| SESR-M5 | TFLITE | float | Qualcomm® SA8775P | 3.052 ms | 0 - 26 MB | NPU
| SESR-M5 | TFLITE | float | Qualcomm® QCS9075 | 3.073 ms | 3 - 9 MB | NPU
| SESR-M5 | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 2.784 ms | 0 - 37 MB | NPU
| SESR-M5 | TFLITE | float | Qualcomm® SA7255P | 10.579 ms | 3 - 26 MB | NPU
| SESR-M5 | TFLITE | float | Qualcomm® SA8295P | 3.481 ms | 3 - 25 MB | NPU
| SESR-M5 | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 1.036 ms | 0 - 24 MB | NPU
| SESR-M5 | TFLITE | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.302 ms | 0 - 26 MB | NPU
| SESR-M5 | TFLITE | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.463 ms | 0 - 31 MB | NPU
| SESR-M5 | TFLITE | w8a8 | Qualcomm® QCS6490 | 2.42 ms | 0 - 3 MB | NPU
| SESR-M5 | TFLITE | w8a8 | Qualcomm® QCS8275 (Proxy) | 1.985 ms | 0 - 23 MB | NPU
| SESR-M5 | TFLITE | w8a8 | Qualcomm® QCS8550 (Proxy) | 0.62 ms | 0 - 1 MB | NPU
| SESR-M5 | TFLITE | w8a8 | Qualcomm® SA8775P | 0.856 ms | 0 - 26 MB | NPU
| SESR-M5 | TFLITE | w8a8 | Qualcomm® QCS9075 | 0.896 ms | 0 - 3 MB | NPU
| SESR-M5 | TFLITE | w8a8 | Qualcomm® QCM6690 | 6.836 ms | 1 - 135 MB | NPU
| SESR-M5 | TFLITE | w8a8 | Qualcomm® QCS8450 (Proxy) | 0.866 ms | 0 - 33 MB | NPU
| SESR-M5 | TFLITE | w8a8 | Qualcomm® SA7255P | 1.985 ms | 0 - 23 MB | NPU
| SESR-M5 | TFLITE | w8a8 | Qualcomm® SA8295P | 1.35 ms | 0 - 21 MB | NPU
| SESR-M5 | TFLITE | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 0.381 ms | 0 - 23 MB | NPU
| SESR-M5 | TFLITE | w8a8 | Snapdragon® 7 Gen 4 Mobile | 0.789 ms | 0 - 22 MB | NPU

## License
* The license for the original implementation of SESR-M5 can be found
  [here](https://github.com/quic/aimet-model-zoo/blob/develop/LICENSE.md).

## References
* [Collapsible Linear Blocks for Super-Efficient Super Resolution](https://arxiv.org/abs/2103.09404)
* [Source Model Implementation](https://github.com/quic/aimet-model-zoo/tree/develop/aimet_zoo_torch/sesr)

## Community
* Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI.
* For questions or feedback please [reach out to us](mailto:ai-hub-support@qti.qualcomm.com).