Add data preprocessing to README.md. Closes #1

README.md

@@ -1,6 +1,6 @@
 # L-CNN — End-to-End Wireframe Parsing
 
-This repository contains the official PyTorch implementation of the paper:  *Yichao Zhou, Haozhi Qi, Yi Ma. ["End-to-End Wireframe Parsing."](https://arxiv.org/abs/1905.03246)  arXiv:1905.03246 [cs.CV]*.
+This repository contains the official PyTorch implementation of the paper:  *[Yichao Zhou](https://yichaozhou.com), [Haozhi Qi](http://haozhi.io), [Yi Ma](https://people.eecs.berkeley.edu/~yima/). ["End-to-End Wireframe Parsing."](https://arxiv.org/abs/1905.03246)  arXiv:1905.03246 [cs.CV]*.
 
 ## Introduction
 
@@ -86,20 +86,35 @@ conda install -y pyyaml docopt matplotlib scikit-image opencv
 mkdir data logs post
 ```
 
-### Downloading data
+### Downloading the Processed Dataset
 Make sure `curl` is installed on your system and execute
 ```bash
 cd data
 ../misc/gdrive-download.sh 1T4_6Nb5r4yAXre3lf-zpmp3RbmyP1t9q wireframe.tar.xz
 tar xf wireframe.tar.xz
-rm *.xz
+rm wireframe.tar.xz
 cd ..
 ```
 
-If `gdrive-download.sh` does not work for you, you can download the pre-processed dataset manually from [Google
+If `gdrive-download.sh` does not work for you, you can download the pre-processed dataset
+`wireframe.tar.xz`  manually from [Google
 Drive](https://drive.google.com/drive/u/1/folders/1rXLAh5VIj8jwf8vLfuZncStihRO2chFr) and proceed
 accordingly.
 
+#### Processing the Dataset
+*Optionally*, you can pre-process (e.g., generate heat maps, do data augmentation) the dataset from
+scratch rather than downloading the processed one.  **Skip** this section if you just want to use
+the pre-processed dataset `wireframe.tar.xz`.
+
+```bash
+cd data
+../misc/gdrive-download.sh 1BRkqyi5CKPQF6IYzj_dQxZFQl0OwbzOf wireframe_raw.tar.xz
+tar xf wireframe_raw.tar.xz
+rm wireframe_raw.tar.xz
+cd ..
+dataset/wireframe.py data/wireframe_raw data/wireframe
+```
+
 ### Training
 To train the neural network on GPU 0 (specified by `-d 0`) with the default parameters, execute
 ```bash

misc/draw-wireframe.py

@@ -1,3 +1,4 @@
+#!/usr/bin/env python3
 import os
 import glob
 import os.path as osp
@@ -10,12 +11,10 @@ import matplotlib.pyplot as plt
 
 from lcnn.utils import parmap
 
-PRED = "/run/media/zyc/YZBackup/Paper/ICCV2019/LCNN/logs/190320-033347-abdf9cb-hg-q2/npz/000072000/*.npz"
-GT = "data/q2/valid/*.npz"
-WF = "/data/wirebase/result/wireframe/wireframe_1_rerun-baseline_0.5_0.5/2/*.mat"
-AFM = "/data/wirebase/result/wireframe/afm/*.npz"
-IMGS = "/data/wirebase/data/v1.1/test/*.jpg"
-LL = "/data/bible/linelet/*.mat"
+GT = "data/wireframe/valid/*.npz"
+WF = "/data/lcnn/wirebase/result/wireframe/wireframe_1_rerun-baseline_0.5_0.5/2/*.mat"
+AFM = "/data/lcnn/wirebase/result/wireframe/afm/*.npz"
+IMGS = "/data/lcnn/wirebase/Wireframe/v1.1/test/*.jpg"
 
 
 def imshow(im):
@@ -34,18 +33,12 @@ def imshow(im):
 
 
 def main():
-    gts = glob.glob(GT)
-    gts.sort()
-    afm = glob.glob(AFM)
-    afm.sort()
-    wf = glob.glob(WF)
-    wf.sort()
-    img = glob.glob(IMGS)
-    img.sort()
-    ll = glob.glob(LL)
-    ll.sort()
-
-    prefix = "/data/wirebase/myplot/"
+    gts = sorted(glob.glob(GT))
+    afm = sorted(glob.glob(AFM))
+    wf = sorted(glob.glob(WF))
+    img = sorted(glob.glob(IMGS))
+
+    prefix = "/data/lcnn/wirebase/myplot/"
     os.makedirs(osp.join(prefix, "GT"), exist_ok=True)
     os.makedirs(osp.join(prefix, "LSD"), exist_ok=True)
     os.makedirs(osp.join(prefix, "AFM"), exist_ok=True)
@@ -53,7 +46,7 @@ def main():
     os.makedirs(osp.join(prefix, "LL"), exist_ok=True)
 
     def draw(args):
-        i, (wf_name, gt_name, afm_name, ll_name, img_name) = args
+        i, (wf_name, gt_name, afm_name, img_name) = args
         img = cv2.imread(img_name, 0)
         lsd = cv2.createLineSegmentDetector(cv2.LSD_REFINE_ADV)
         lsd_line, _, _, lsd_score = lsd.detect(img)
@@ -72,25 +65,13 @@ def main():
         wf_line = scipy.io.loadmat(wf_name)["lines"].reshape(-1, 2, 2)
         wf_line = wf_line[:, :, ::-1]
 
-        ll_line = scipy.io.loadmat(wf_name)["lines"].reshape(-1, 2, 2)
-        ll_line = ll_line[:, :, ::-1]
-
-        plt.figure("LL")
-        imshow(img)
-        for a, b in ll_line - 0.5:
-            plt.plot([a[1], b[1]], [a[0], b[0]], color="orange", linewidth=0.5)
-            plt.scatter(a[1], a[0], color="#33FFFF", s=1.2, edgecolors="none", zorder=5)
-            plt.scatter(b[1], b[0], color="#33FFFF", s=1.2, edgecolors="none", zorder=5)
-        plt.savefig(osp.join(prefix, "LL", f"{i:05}"), dpi=500, bbox_inches=0)
-        plt.close()
-
         plt.figure("GT")
         imshow(img)
         for a, b in gt_line - 0.5:
             plt.plot([a[1], b[1]], [a[0], b[0]], color="orange", linewidth=0.5)
             plt.scatter(a[1], a[0], color="#33FFFF", s=1.2, edgecolors="none", zorder=5)
             plt.scatter(b[1], b[0], color="#33FFFF", s=1.2, edgecolors="none", zorder=5)
-        plt.savefig(osp.join(prefix, "GT", f"{i:05}"), dpi=500, bbox_inches=0)
+        plt.savefig(osp.join(prefix, "GT", f"{i:05}"), dpi=3000, bbox_inches=0)
         plt.close()
 
         plt.figure("LSD")
@@ -120,7 +101,7 @@ def main():
         plt.savefig(osp.join(prefix, "WF", f"{i:05}"), dpi=3000, bbox_inches=0)
         plt.close()
 
-    parmap(draw, enumerate(zip(wf, gts, afm, ll, img)))
+    parmap(draw, enumerate(zip(wf, gts, afm, img)))
 
 
 if __name__ == "__main__":