Merge pull request #2954 from Wovchena/models/intel/handwritten-english-recognition-0001-fix-style

models/intel/handwritten-english-recognition-0001 fix style

Author: Wovchena

ci/check-basics.py

@@ -51,7 +51,7 @@ def complain(message):
 
     print('running miscellaneous checks...', flush=True)
 
-    if subprocess.run(['git', '--no-pager', 'diff', '--check', empty_tree_hash, '--'],
+    if subprocess.run(['git', '--no-pager', 'diff', '--check', empty_tree_hash, '--', ':(exclude)data/dataset_classes/gnhk.txt'],
             cwd=OMZ_ROOT).returncode != 0:
         all_passed = False
 

demos/handwritten_text_recognition_demo/python/handwritten_text_recognition_demo.py

@@ -91,7 +91,7 @@ def main():
         out_blob = args.output_blob
     else:
         assert len(net.outputs) == 1, "Demo supports only single output topologies"
-        out_blob = next(iter(net.output_info))
+        out_blob = next(iter(net.outputs))
 
     characters = get_characters(args)
     codec = CTCCodec(characters, args.designated_characters, args.top_k)

demos/handwritten_text_recognition_demo/python/models.lst

@@ -1,4 +1,5 @@
 # This file can be used with the --list option of the model downloader.
-handwritten-japanese-recognition-0001
-handwritten-simplified-chinese-recognition-0001
-handwritten-english-recognition-0001
+handwritten-japanese-recognition-????
+handwritten-simplified-chinese-recognition-????
+# TODO
+# handwritten-english-recognition-????

models/intel/device_support.md

@@ -34,6 +34,7 @@
 | formula-recognition-polynomials-handwritten-0001-decoder | YES | YES |    |
 | formula-recognition-polynomials-handwritten-0001-encoder | YES | YES |    |
 | gaze-estimation-adas-0002 | YES | YES | YES |
+| handwritten-english-recognition-0001 | YES | YES | YES |
 | handwritten-japanese-recognition-0001 | YES | YES | YES |
 | handwritten-score-recognition-0003 | YES | YES | YES |
 | handwritten-simplified-chinese-recognition-0001 | YES | YES | YES |

models/intel/handwritten-english-recognition-0001/README.md

@@ -18,7 +18,7 @@ The network is able to recognize English text consisting of characters in the [G
 | Accuracy on GNHK test subset (excluding images wider than 2000px after resized to height 96px with aspect ratio) | 81.5%     |
 | Source framework          | PyTorch\* |
 
-> **Note:** to achieve the accuracy, images from the GNHK test set should be binarized using adaptive thresholding, and preprocessed into single-line text images, using the coordinates from the accompanying JSON annotation files in the GNHK dataset. See [preprocess_gnhk.py](./preprocess_gnhk.py).
+> **Note:** to achieve the accuracy, images from the GNHK test set should be binarized using adaptive thresholding, and preprocessed into single-line text images, using the coordinates from the accompanying JSON annotation files in the GNHK dataset. See `<omz_dir>/models/intel/handwritten-english-recognition-0001/preprocess_gnhk.py`.
 
 This model adopts [label error rate](https://dl.acm.org/doi/abs/10.1145/1143844.1143891) as the metric for accuracy.
 

models/intel/handwritten-english-recognition-0001/accuracy-check.yml

@@ -28,4 +28,3 @@ models:
         metrics:
           - type: label_level_recognition_accuracy
             reference: 0.815
-

models/intel/handwritten-english-recognition-0001/preprocess_gnhk.py

@@ -56,7 +56,7 @@ def main():
             line_indices = set(map(lambda obj: obj["line_idx"], data))
             img = cv2.imread(image, cv2.IMREAD_GRAYSCALE)
             img = binarize(img)
-            for idx in sorted(list(line_indices)):
+            for idx in sorted(line_indices):
                 objects = list(filter(lambda obj: obj["line_idx"] == idx, data))
                 # discard math symbols, scribbles, illegible text, and printed text
                 objects = list(filter(lambda obj: obj["text"] != "%math%" and obj["text"] != "%SC%" and obj["text"] != "%NA%" and obj["type"] != "P", objects))
@@ -69,26 +69,26 @@ def main():
                 # create mask for the words
                 mask = np.zeros(img.shape[0:2], dtype=np.uint8)
                 for obj in objects:
-                        region = [
-                            [obj["polygon"]["x0"], obj["polygon"]["y0"]],
-                            [obj["polygon"]["x1"], obj["polygon"]["y1"]],
-                            [obj["polygon"]["x2"], obj["polygon"]["y2"]],
-                            [obj["polygon"]["x3"], obj["polygon"]["y3"]]
-                        ]
-                        points = np.array([region])
-                        cv2.drawContours(mask, [points], -1, (255, 255, 255), -1, cv2.LINE_AA)
+                    region = [
+                        [obj["polygon"]["x0"], obj["polygon"]["y0"]],
+                        [obj["polygon"]["x1"], obj["polygon"]["y1"]],
+                        [obj["polygon"]["x2"], obj["polygon"]["y2"]],
+                        [obj["polygon"]["x3"], obj["polygon"]["y3"]]
+                    ]
+                    points = np.array([region])
+                    cv2.drawContours(mask, [points], -1, (255, 255, 255), -1, cv2.LINE_AA)
                 masked = cv2.bitwise_and(img, img, mask = mask)
                 bg = np.ones_like(img, np.uint8) * 255
                 cv2.bitwise_not(bg, bg, mask = mask)
                 overlay = bg + masked
                 # crop bounding rectangle of the text region
-                l = list(map(lambda obj: [
+                polys = list(map(lambda obj: [
                     [obj["polygon"]["x0"], obj["polygon"]["y0"]],
                     [obj["polygon"]["x1"], obj["polygon"]["y1"]],
                     [obj["polygon"]["x2"], obj["polygon"]["y2"]],
                     [obj["polygon"]["x3"], obj["polygon"]["y3"]]
-                    ], objects))
-                flat = [item for sublist in l for item in sublist]
+                ], objects))
+                flat = [item for sublist in polys for item in sublist]
                 pts = np.array(flat)
                 rect = cv2.boundingRect(pts)
                 x, y, w, h = rect
@@ -107,4 +107,3 @@ def main():
 
 if __name__ == '__main__':
     main()
-