update readme

README.md

@@ -4,13 +4,12 @@
 ## deepCR: Deep Learning Based Cosmic Ray Removal for Astronomical Images
 
 Identify and remove cosmic rays from astronomical images using trained convolutional neural networks.
-Fast on both CPU and GPU.
 
-This is the installable package which implements the methods described in the paper: Zhang & Bloom (2019), submitted to AAS Journals.
+This is the installable package which implements the methods described in the paper: Zhang & Bloom (2019), submitted.
 
 Code to reproduce benchmarking results in the paper is at: https://github.com/kmzzhang/deepCR-paper
 
-If you use this package, please cite Zhang & Bloom (2019): www.arxiv.org/XXX
+If you use this package, please cite Zhang & Bloom (2019): url TBA
 
 <img src="https://raw.githubusercontent.com/profjsb/deepCR/master/imgs/postage-sm.jpg" wdith="90%">
 
@@ -39,11 +38,13 @@ image = fits.getdata("example_flc.fits")
 
 # create an instance of deepCR with specified model configuration
 mdl = deepCR(mask="ACS-WFC-F606W-2-32",
-	         inpaint="ACS-WFC-F606W-3-32",
+	     inpaint="ACS-WFC-F606W-3-32",
              device="CPU")
+
 # apply to input image
 mask, cleaned_image = mdl.clean(image, threshold = 0.5)
-# examine those outputs to choose an adequate threshold
+# visualize those outputs to choose an adequate threshold
+# note that deepCR-inpaint would be inaccurate if mask does not fully cover CR
 
 # if you only need CR mask you may skip image inpainting for shorter runtime
 mask = mdl.clean(image, threshold = 0.5, inpaint=False)
@@ -52,14 +53,11 @@ mask = mdl.clean(image, threshold = 0.5, inpaint=False)
 prob_mask = mdl.clean(image, binary=False)
 ```
 
-To reduce memory consumption (recommended for image larger than 1k x 1k), you can tell deepCR to segment input image to smaller patches of seg * seg.
+To reduce memory consumption (recommended for images larger 0.5 Mpix), you can tell deepCR to segment the input image into 256*256 patches, and process one patch at a time.
 ```python
-mask, cleaned_image = mdl.clean(image, threshold = 0.5, seg = 256)
-mask = mdl.clean(image, threshold = 0.5,  seg = 256)
+mask, cleaned_image = mdl.clean(image, threshold = 0.5, seg = True)
+mask = mdl.clean(image, threshold = 0.5, seg = True)
 ```
-deepCR perform CR rejection on small patches one at a time. It then stitches the predictions back to the original image size.
-
-We recommend using patch size (seg) no smaller than 64. seg is by default 256.
 
 ### Currently available models
 
@@ -75,14 +73,14 @@ inpaint:
 
     ACS-WFC-F606W-3-32(*)
 
-Recommended models are marked in (*).
+Recommended models are marked in (*). Larger number indicate larger capacity and better performance.
 
 Input images should come from _flc.fits files which are in units of electrons.
-Larger number indicate larger capacity and better performance.
+
 
 ### API Documentation
 
-Documentation is under development at: https://deepcr.readthedocs.io/en/latest/deepCR.html
+Full documentation is under development at: https://deepcr.readthedocs.io/en/latest/deepCR.html
 
 ### Limitations and Caveats