Update EMT.py
Browse files
EMT.py
CHANGED
@@ -67,62 +67,72 @@ class EMT(datasets.GeneratorBasedBuilder):
|
|
67 |
|
68 |
def _split_generators(self, dl_manager):
|
69 |
"""Download train/test images and annotations."""
|
70 |
-
|
71 |
-
|
72 |
-
|
73 |
-
"train": dl_manager.download_and_extract(_TRAIN_IMAGE_ARCHIVE_URL),
|
74 |
-
"test": dl_manager.download_and_extract(_TEST_IMAGE_ARCHIVE_URL),
|
75 |
}
|
76 |
-
|
77 |
-
# Download
|
78 |
-
|
79 |
-
"train":
|
80 |
-
"test":
|
81 |
}
|
82 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
83 |
return [
|
84 |
datasets.SplitGenerator(
|
85 |
name=datasets.Split.TRAIN,
|
86 |
gen_kwargs={
|
87 |
-
"images":
|
88 |
-
"annotation_path":
|
89 |
},
|
90 |
),
|
91 |
datasets.SplitGenerator(
|
92 |
name=datasets.Split.TEST,
|
93 |
gen_kwargs={
|
94 |
-
"images":
|
95 |
-
"annotation_path":
|
96 |
},
|
97 |
),
|
98 |
]
|
99 |
-
|
100 |
def _generate_examples(self, images, annotation_path):
|
101 |
"""Generate dataset examples by matching images to their corresponding annotations."""
|
102 |
|
103 |
-
# Load ALL annotations into memory before iterating over images
|
104 |
annotations = {}
|
105 |
-
|
106 |
-
|
|
|
|
|
107 |
ann_path = os.path.join(annotation_path, ann_file)
|
108 |
-
|
109 |
-
|
110 |
with open(ann_path, "r", encoding="utf-8") as f:
|
111 |
for line in f:
|
112 |
parts = line.strip().split()
|
113 |
if len(parts) < 8:
|
114 |
continue
|
115 |
-
|
116 |
frame_id, track_id, class_name = parts[:3]
|
117 |
bbox = list(map(float, parts[4:8]))
|
118 |
class_id = _GT_OBJECT_CLASSES.get(class_name, -1)
|
119 |
img_name = f"{frame_id}.jpg"
|
120 |
-
|
121 |
-
# Store annotation in a
|
122 |
key = f"{video_name}/{img_name}"
|
123 |
if key not in annotations:
|
124 |
annotations[key] = []
|
125 |
-
|
126 |
annotations[key].append(
|
127 |
{
|
128 |
"bbox": bbox,
|
@@ -131,14 +141,14 @@ class EMT(datasets.GeneratorBasedBuilder):
|
|
131 |
"class_name": class_name,
|
132 |
}
|
133 |
)
|
134 |
-
|
135 |
# Yield dataset entries
|
136 |
idx = 0
|
137 |
for file_path, file_obj in images:
|
138 |
img_name = os.path.basename(file_path)
|
139 |
-
video_name = os.path.basename(os.path.dirname(file_path))
|
140 |
-
key = f"{video_name}/{img_name}"
|
141 |
-
|
142 |
if key in annotations:
|
143 |
yield idx, {
|
144 |
"image": {"path": file_path, "bytes": file_obj.read()},
|
|
|
67 |
|
68 |
def _split_generators(self, dl_manager):
|
69 |
"""Download train/test images and annotations."""
|
70 |
+
image_urls = {
|
71 |
+
"train": _TRAIN_IMAGE_ARCHIVE_URL,
|
72 |
+
"test": _TEST_IMAGE_ARCHIVE_URL,
|
|
|
|
|
73 |
}
|
74 |
+
|
75 |
+
# Download the individual annotation files for train and test
|
76 |
+
annotation_urls = {
|
77 |
+
"train": _ANNOTATION_REPO + "/train/",
|
78 |
+
"test": _ANNOTATION_REPO + "/test/",
|
79 |
}
|
80 |
+
|
81 |
+
# Download image files
|
82 |
+
images = {
|
83 |
+
"train": dl_manager.iter_archive(image_urls["train"]),
|
84 |
+
"test": dl_manager.iter_archive(image_urls["test"]),
|
85 |
+
}
|
86 |
+
|
87 |
+
# Download annotation files
|
88 |
+
annotations = {
|
89 |
+
"train": dl_manager.download_and_extract(annotation_urls["train"]),
|
90 |
+
"test": dl_manager.download_and_extract(annotation_urls["test"]),
|
91 |
+
}
|
92 |
+
|
93 |
return [
|
94 |
datasets.SplitGenerator(
|
95 |
name=datasets.Split.TRAIN,
|
96 |
gen_kwargs={
|
97 |
+
"images": images["train"],
|
98 |
+
"annotation_path": annotations["train"],
|
99 |
},
|
100 |
),
|
101 |
datasets.SplitGenerator(
|
102 |
name=datasets.Split.TEST,
|
103 |
gen_kwargs={
|
104 |
+
"images": images["test"],
|
105 |
+
"annotation_path": annotations["test"],
|
106 |
},
|
107 |
),
|
108 |
]
|
109 |
+
|
110 |
def _generate_examples(self, images, annotation_path):
|
111 |
"""Generate dataset examples by matching images to their corresponding annotations."""
|
112 |
|
|
|
113 |
annotations = {}
|
114 |
+
|
115 |
+
# Load all annotations into memory
|
116 |
+
for ann_file in os.listdir(annotation_path):
|
117 |
+
video_name = os.path.splitext(ann_file)[0] # Get video folder name
|
118 |
ann_path = os.path.join(annotation_path, ann_file)
|
119 |
+
|
|
|
120 |
with open(ann_path, "r", encoding="utf-8") as f:
|
121 |
for line in f:
|
122 |
parts = line.strip().split()
|
123 |
if len(parts) < 8:
|
124 |
continue
|
125 |
+
|
126 |
frame_id, track_id, class_name = parts[:3]
|
127 |
bbox = list(map(float, parts[4:8]))
|
128 |
class_id = _GT_OBJECT_CLASSES.get(class_name, -1)
|
129 |
img_name = f"{frame_id}.jpg"
|
130 |
+
|
131 |
+
# Store annotation in a dictionary
|
132 |
key = f"{video_name}/{img_name}"
|
133 |
if key not in annotations:
|
134 |
annotations[key] = []
|
135 |
+
|
136 |
annotations[key].append(
|
137 |
{
|
138 |
"bbox": bbox,
|
|
|
141 |
"class_name": class_name,
|
142 |
}
|
143 |
)
|
144 |
+
|
145 |
# Yield dataset entries
|
146 |
idx = 0
|
147 |
for file_path, file_obj in images:
|
148 |
img_name = os.path.basename(file_path)
|
149 |
+
video_name = os.path.basename(os.path.dirname(file_path)) # Match the video folder
|
150 |
+
key = f"{video_name}/{img_name}"
|
151 |
+
|
152 |
if key in annotations:
|
153 |
yield idx, {
|
154 |
"image": {"path": file_path, "bytes": file_obj.read()},
|