Update README.md

README.md

@@ -7,15 +7,15 @@ license: apache-2.0
 
 # Ultra Zoom
 
-A fast single image super-resolution (SISR) model for upscaling images without loss of detail. Ultra Zoom uses a two-stage "zoom in and enhance" strategy that uses a fast deterministic upscaling algorithm to zoom in and then enhances the image through a residual pathway that operates primarily in the low-resolution subspace of a deep neural network. As such, Ultra Zoom requires less resources than upscalers that  necessarily predict every new pixel de novo - making it outstanding for real-time image processing.
+A fast single image super-resolution (SISR) model for upscaling images without loss of detail. Ultra Zoom uses a two-stage "zoom in and enhance" strategy that uses a fast deterministic upscaling algorithm to zoom in and then enhances the image through a residual pathway that operates primarily in the low-resolution subspace of a deep neural network. As such, Ultra Zoom requires less resources than upscalers that predict every new pixel de novo - making it outstanding for real-time image processing.
 
-- **Fast and scalable**: Instead of predicting the individual pixels of the upscaled image, Ultra Zoom uses a unique "zoom in and enhance" approach that combines the speed of bicubic interpolation with the power of a deep neural network.
+## Key Features
 
-- **Full RGB**: Unlike many efficient SR models that only operate in the luminance domain, Ultra Zoom operates within the full RGB color spectrum enhancing both luminance and chrominance for the best possible quality.
+- **Fast and scalable**: Instead of predicting the individual pixels of the upscaled image, Ultra Zoom uses a unique "zoom in and enhance" approach that combines the speed of deterministic bicubic interpolation with the power of a deep neural network.
 
-- **Denoising and Deblurring**: During the enhancement stage, the model removes multiple types of noise and blur making images look crisp and clean.
+- **Full RGB**: Unlike many efficient SR models that only operate in the luminance domain, Ultra Zoom operates within the full RGB color domain enhancing both luminance and chrominance for the best possible quality.
 
-- **Next-gen architecture**: Ultra Zoom employs a convolutional neural network architecture that performs better than previous generations by employing a modern architecture with spatial attention, wide non-linear activations, and sub-pixel convolution.
+- **Denoising and Deblurring**: During the enhancement stage, the model removes multiple types of noise and blur making images look crisp and clean.
 
 ## Pretrained Models
 
@@ -24,8 +24,8 @@ The following pretrained models are available on HuggingFace Hub.
 | Name | Zoom | Num Channels | Hidden Ratio | Encoder Layers | Total Parameters |
 |---|---|---|---|---|---|
 | [andrewdalpino/UltraZoom-2X](https://huggingface.co/andrewdalpino/UltraZoom-2X) | 2X | 48 | 2X | 20 | 1.8M |
-| [andrewdalpino/UltraZoom-4X](https://huggingface.co/andrewdalpino/UltraZoom-4X) | 4X | 96 | 2X | 28 | 10M |
-| [andrewdalpino/UltraZoom-8X](https://huggingface.co/andrewdalpino/UltraZoom-8X) | 8X | 192 | 2X | 36 | 54M |
+| [andrewdalpino/UltraZoom-3X](https://huggingface.co/andrewdalpino/UltraZoom-3X) | 3X | 54 | 2X | 30 | 3.5M |
+| [andrewdalpino/UltraZoom-4X](https://huggingface.co/andrewdalpino/UltraZoom-4X) | 4X | 96 | 2X | 40 | 14M |
 
 ## Pretrained Example