|
|
--- |
|
|
license: other |
|
|
license_name: aimet-model-zoo-license |
|
|
license_link: https://github.com/quic/aimet-model-zoo/blob/develop/LICENSE.pdf |
|
|
pipeline_tag: image-to-image |
|
|
tags: |
|
|
- AIoT |
|
|
- QNN |
|
|
--- |
|
|
|
|
|
 |
|
|
|
|
|
## XLSR (Extreme Lightweight Super Resolution): Super Resolution |
|
|
|
|
|
XLSR is an ultra-lightweight image super-resolution model designed for mobile devices, featuring extremely low parameter count and model size. It supports quantization acceleration to enable efficient image upscaling and enhancement in resource-constrained environments. With hardware-aware design and Clipped ReLU activation, XLSR maintains strong reconstruction quality after quantization, making it suitable for real-time video processing and on-device image enhancement. |
|
|
|
|
|
### Source model |
|
|
|
|
|
- Input shape: 1x3x128x128 |
|
|
- Number of parameters: 0.03M |
|
|
- Model size: 172KB |
|
|
- Output shape: 1x3x512x512 |
|
|
|
|
|
The source model can be found [here](https://github.com/quic/aimet-model-zoo/tree/develop/aimet_zoo_torch/xlsr) |
|
|
|
|
|
## Performance Reference |
|
|
|
|
|
Please search model by model name in [Model Farm](https://aiot.aidlux.com/en/models) |
|
|
|
|
|
## Inference & Model Conversion |
|
|
|
|
|
Please search model by model name in [Model Farm](https://aiot.aidlux.com/en/models) |
|
|
|
|
|
## License |
|
|
|
|
|
- Source Model: [AIMET-MODEL-ZOO](https://github.com/quic/aimet-model-zoo/blob/develop/LICENSE.pdf) |
|
|
|
|
|
- Deployable Model: [AIMET-MODEL-ZOO](https://github.com/quic/aimet-model-zoo/blob/develop/LICENSE.pdf) |
