Merge branch 'main' into aj/sam2.1

# Conflicts:
#	sam2/modeling/sam2_base.py
#	sam2/sam2_video_predictor.py
#	setup.py

sam2/utils/misc.py

@@ -346,6 +346,31 @@ def load_video_frames_from_memory(
     return images, video_height, video_width
 
 
+def load_video_frames_from_memory(
+    imgs,
+    image_size,
+    offload_video_to_cpu,
+    img_mean=(0.485, 0.456, 0.406),
+    img_std=(0.229, 0.224, 0.225),
+):
+    img_mean = torch.tensor(img_mean, dtype=torch.float32)[:, None, None]
+    img_std = torch.tensor(img_std, dtype=torch.float32)[:, None, None]
+
+    images = torch.zeros(len(imgs), 3, image_size, image_size, dtype=torch.float32)
+
+    for n, img in enumerate(imgs):
+        images[n], video_height, video_width = _load_img_np_as_tensor(img, image_size)
+
+    if not offload_video_to_cpu:
+        images = images.cuda()
+        img_mean = img_mean.cuda()
+        img_std = img_std.cuda()
+
+    images -= img_mean
+    images /= img_std
+    return images, video_height, video_width
+
+
 def fill_holes_in_mask_scores(mask, max_area):
     """
     A post processor to fill small holes in mask scores with area under `max_area`.