Update README.md

README.md

@@ -26,10 +26,14 @@ SuperCool is a fast single-image super-resolution (SISR) model capable of upscal
 
 - **Very fast inference**: Instead of directly predicting the individual pixels of the upscaled image, SuperCool uses a fast deterministic  upscaling algorithm and then fills in the missing details through a residual pathway that operates primarily within the low-resolution subspace. As such, the model is capable of being used for real-time image processing.
 
-- **Train on your own images**: SuperCool's dataloader and preprocessing pipeline works with any of the popular image formats including JPEG, PNG, WEBP, and GIF. Just point the training script to the location of the folder containing your training images and begin training. This enabled you to build upscalers that specialize on certain types of images such as satellite photos or portraits.
+- **Train on your own images**: SuperCool's dataloader and preprocessing pipeline works with any of the popular image formats including JPEG, PNG, WEBP, and GIF. Just point the training script to the location of the folder containing your training images and begin training. This enables you to build upscalers that specialize on certain types of images such as satellite photos or portraits.
 
 - **Adjustable model size**: Depending on your computing and memory budget, you can train larger or smaller models by adjusting a few hyper-parameters such as the number of hidden channels, number of hidden layers, and size/strength of the adversarial model.
 
+## Comparison
+
+![Super Cool 4X Small Comparison](https://raw.githubusercontent.com/andrewdalpino/SuperCool/master/docs/images/comparison-4x-small.png)
+
 ## Install Project Dependencies
 
 Project dependencies are specified in the `requirements.txt` file. You can install them with [pip](https://pip.pypa.io/en/stable/) using the following command from the project root. We recommend using a virtual environment such as `venv` to keep package dependencies on your system tidy.
@@ -191,4 +195,4 @@ python upscale.py --checkpoint_path="./checkpoints/fine-tuned.pt" --image_path="
 >- W. Shi, et al. Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network, 2016.
 >- T. Salimans, et al. Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks, OpenAI, 2016.
 >- C. Dong, et al. Image Super-Resolution Using Deep Convolutional Networks, 2015.
->- J. Kim, et a. Accurate Image Super-Resolution Using Very Deep Convolutional Networks.
+>- J. Kim, et a. Accurate Image Super-Resolution Using Very Deep Convolutional Networks.