chore(data): save memory for COCO dataset (Megvii-BaseDetection#1066)

yolox/data/datasets/coco.py

@@ -13,6 +13,24 @@
 from .datasets_wrapper import Dataset
 
 
+def remove_useless_info(coco):
+    """
+    Remove useless info in coco dataset. COCO object is modified inplace.
+    This function is mainly used for saving memory (save about 30% mem).
+    """
+    if isinstance(coco, COCO):
+        dataset = coco.dataset
+        dataset.pop("info", None)
+        dataset.pop("licenses", None)
+        for img in dataset["images"]:
+            img.pop("license", None)
+            img.pop("coco_url", None)
+            img.pop("date_captured", None)
+            img.pop("flickr_url", None)
+        for anno in coco.dataset["annotations"]:
+            anno.pop("segmentation", None)
+
+
 class COCODataset(Dataset):
     """
     COCO dataset class.
@@ -43,6 +61,7 @@ def __init__(
         self.json_file = json_file
 
         self.coco = COCO(os.path.join(self.data_dir, "annotations", self.json_file))
+        remove_useless_info(self.coco)
         self.ids = self.coco.getImgIds()
         self.class_ids = sorted(self.coco.getCatIds())
         cats = self.coco.loadCats(self.coco.getCatIds())