flushing out readme

README.md

@@ -5,6 +5,25 @@
 |------------------------------------------------------------------------------------------------------------|--------------------------------------------------------------------------------------------------------------------|
 | [![PaperVideo](https://img.youtube.com/vi/RvQIZ7aNOc0/0.jpg)](https://www.youtube.com/watch?v=RvQIZ7aNOc0) | [![PosterPDF](./materials/Mini-Meta-Figure.png)](https://drive.google.com/file/d/1g1XR3ZRRs6CCAxeiSVdE6fcffyXVxdOh/view?usp=sharing) |
 
+## Directory Structure
+    .
+    ├── configs                 # Stores yaml files for configuring training, evaluation, demo, etc.
+    ├── data                    # Directory for data. See data.md for details on getting data.
+    ├── materials               # Miscellaneous assets for the report, poster, video, etc.
+    ├── scene_demo              # Small sample of input data used to demo adaptive forward-pass.
+    ├── src                     # Jupyter notebook series explaining how apply radar concepts to real data.
+    `── .
+        ├── core                # Contains scripts used for configuration as well as the loss metrics used.
+        ├── dataset             # Contains scripts used to read in the data, a data generator which will sample batches of data, as well as an hdr visualization script.
+        ├── model               # Contains scripts which define the UNet for the inner model, as well as functions for training, validating, and evaluating our model with MAML.
+    ├── scripts                 # Various setup scripts for mmwavestudio, etc
+    ├── .gitignore
+    ├── README.md
+    ├── demo.py                 # Runs an adaptive forward pass to demonstrate deployed usage.
+    ├── eval.py                 # Runs an evaluation of a trained model.
+    ├── requirements.txt        # Required dependencies to run this model.
+    └── train.py                # Runs training iterations.
+
 ## Getting Started
 MetaHDR was implemented and tested on CentOS 7 with python >=3.9.0, and supports GPU and CPU computation. We have also provided a Colab demo for convenience.
 

demo.py

@@ -58,6 +58,7 @@ def main(args):
     HDR_inputs = np.array(HDR_inputs)
 
     if crop_flag:
+        print("Cropping images about center...")
         IMG_HEIGHT = LDR_inputs.shape[1]
         IMG_WIDTH = LDR_inputs.shape[2]
         crop_factor = 0.5
@@ -116,7 +117,7 @@ def main(args):
         HDR_reconst, test_ssim, test_psnr = evaluate_maml(meta_model, loss_func, training, testing, 0, cfg.EVAL.NUM_TASK_TR_ITER, device=device, model_type=cfg.TRAIN.MODEL, visualize_flag=True, visualize_dir=output_dir)
 
     # Save gamma corrected output image
-    io.imsave(f"{output_dir}/HDR{0:03d}.png", visualize_hdr_image(HDR_reconst))
+    io.imsave(f"{output_dir}/HDR{0:03d}.png", visualize_hdr_image(HDR_reconst).astype(np.uint8))
 
     return