Fix in ImageNet_txt (#302)

* Add extensions for images to annotation file

* Remove image search in extractor

* Update changelog

Co-authored-by: Maxim Zhiltsov <[email protected]>

CHANGELOG.md

@@ -18,6 +18,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Changed
 - Tensorflow AVX check is made optional in API and is disabled by default (<https://github.com/openvinotoolkit/datumaro/pull/305>)
+- Extensions for images in ImageNet_txt are now mandatory (<https://github.com/openvinotoolkit/datumaro/pull/302>)
 
 ### Deprecated
 -

datumaro/plugins/imagenet_txt_format.py

@@ -10,7 +10,6 @@
     LabelCategories, AnnotationType, SourceExtractor, Importer
 )
 from datumaro.components.converter import Converter
-from datumaro.util.image import find_images
 
 
 class ImagenetTxtPath:
@@ -49,26 +48,22 @@ def _load_categories(self, labels):
     def _load_items(self, path):
         items = {}
 
-        image_dir = self.image_dir
-        if osp.isdir(image_dir):
-            images = { osp.splitext(osp.relpath(p, image_dir))[0]: p
-                for p in find_images(image_dir, recursive=True) }
-        else:
-            images = {}
-
         with open(path, encoding='utf-8') as f:
             for line in f:
                 item = line.split('\"')
                 if 1 < len(item):
                     if len(item) == 3:
                         item_id = item[1]
-                        label_ids = [int(id) for id in item[2].split()]
+                        item = item[2].split()
+                        image = item_id + item[0]
+                        label_ids = [int(id) for id in item[1:]]
                     else:
                         raise Exception("Line %s: unexpected number "
                             "of quotes in filename" % line)
                 else:
                     item = line.split()
-                    item_id = item[0]
+                    item_id = osp.splitext(item[0])[0]
+                    image = item[0]
                     label_ids = [int(id) for id in item[1:]]
 
                 anno = []
@@ -79,7 +74,7 @@ def _load_items(self, path):
                     anno.append(Label(label))
 
                 items[item_id] = DatasetItem(id=item_id, subset=self._subset,
-                    image=images.get(item_id), annotations=anno)
+                    image=osp.join(self.image_dir, image), annotations=anno)
 
         return items
 
@@ -105,18 +100,20 @@ def apply(self):
 
             labels = {}
             for item in subset:
-                labels[item.id] = set(p.label for p in item.annotations
+                item_id = item.id
+                if 1 < len(item_id.split()):
+                    item_id = '\"' + item_id + '\"'
+                item_id += self._find_image_ext(item)
+                labels[item_id] = set(p.label for p in item.annotations
                     if p.type == AnnotationType.label)
 
                 if self._save_images and item.has_image:
                     self._save_image(item, subdir=ImagenetTxtPath.IMAGE_DIR)
 
             annotation = ''
             for item_id, item_labels in labels.items():
-                if 1 < len(item_id.split()):
-                    item_id = '\"' + item_id + '\"'
-                annotation += '%s %s\n' % (
-                    item_id, ' '.join(str(l) for l in item_labels))
+                annotation += '%s %s\n' % (item_id,
+                    ' '.join(str(l) for l in item_labels))
 
             with open(annotation_file, 'w', encoding='utf-8') as f:
                 f.write(annotation)

tests/assets/imagenet_txt_dataset/train.txt

@@ -1,4 +1,4 @@
-1 0
-2 5
-3 3
-4 5
+1.jpg 0
+2.jpg 5
+3.jpg 3
+4.jpg 5