iicd/cotracker
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Merge pull request #14 from JunkyByte/main

Browse Source 
minor fixes / mps default device when available / occlusion visualization
This commit is contained in:
Nikita Karaev 2023-07-27 13:36:01 +01:00 committed by GitHub
commit e84ca71ba5
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
3 changed files with 12 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
cotracker/predictor.py
View File

@ -133,7 +133,7 @@ class CoTrackerPredictor(torch.nn.Module):
) )
if add_support_grid: if add_support_grid:
grid_pts = get_points_on_a_grid(self.support_grid_size, self.interp_shape) grid_pts = get_points_on_a_grid(self.support_grid_size, self.interp_shape, device=video.device)
grid_pts = torch.cat( grid_pts = torch.cat(
[torch.zeros_like(grid_pts[:, :, :1]), grid_pts], dim=2 [torch.zeros_like(grid_pts[:, :, :1]), grid_pts], dim=2
) )

5
cotracker/utils/visualizer.py
View File

@ -62,6 +62,7 @@ class Visualizer:
self, self,
video: torch.Tensor, # (B,T,C,H,W) video: torch.Tensor, # (B,T,C,H,W)
tracks: torch.Tensor, # (B,T,N,2) tracks: torch.Tensor, # (B,T,N,2)
visibility: torch.Tensor = None, # (B, T, N, 1) bool
gt_tracks: torch.Tensor = None, # (B,T,N,2) gt_tracks: torch.Tensor = None, # (B,T,N,2)
segm_mask: torch.Tensor = None, # (B,1,H,W) segm_mask: torch.Tensor = None, # (B,1,H,W)
filename: str = "video", filename: str = "video",
@ -93,6 +94,7 @@ class Visualizer:
res_video = self.draw_tracks_on_video( res_video = self.draw_tracks_on_video(
video=video, video=video,
tracks=tracks, tracks=tracks,
visibility=visibility,
segm_mask=segm_mask, segm_mask=segm_mask,
gt_tracks=gt_tracks, gt_tracks=gt_tracks,
query_frame=query_frame, query_frame=query_frame,
@ -126,6 +128,7 @@ class Visualizer:
self, self,
video: torch.Tensor, video: torch.Tensor,
tracks: torch.Tensor, tracks: torch.Tensor,
visibility: torch.Tensor = None,
segm_mask: torch.Tensor = None, segm_mask: torch.Tensor = None,
gt_tracks=None, gt_tracks=None,
query_frame: int = 0, query_frame: int = 0,
@ -227,11 +230,13 @@ class Visualizer:
if not compensate_for_camera_motion or ( if not compensate_for_camera_motion or (
compensate_for_camera_motion and segm_mask[i] > 0 compensate_for_camera_motion and segm_mask[i] > 0
): ):
cv2.circle( cv2.circle(
res_video[t], res_video[t],
coord, coord,
int(self.linewidth * 2), int(self.linewidth * 2),
vector_colors[t, i].tolist(), vector_colors[t, i].tolist(),
thickness=-1 if visibility[0, t, i] else 2
-1, -1,
) )

12
demo.py
View File

@ -14,6 +14,9 @@ from PIL import Image
from cotracker.utils.visualizer import Visualizer, read_video_from_path from cotracker.utils.visualizer import Visualizer, read_video_from_path
from cotracker.predictor import CoTrackerPredictor from cotracker.predictor import CoTrackerPredictor
DEFAULT_DEVICE = ('cuda' if torch.cuda.is_available() else
'mps' if torch.backends.mps.is_available() else
'cpu')
if __name__ == "__main__": if __name__ == "__main__":
parser = argparse.ArgumentParser() parser = argparse.ArgumentParser()
@ -55,11 +58,8 @@ if __name__ == "__main__":
segm_mask = torch.from_numpy(segm_mask)[None, None] segm_mask = torch.from_numpy(segm_mask)[None, None]
model = CoTrackerPredictor(checkpoint=args.checkpoint) model = CoTrackerPredictor(checkpoint=args.checkpoint)
if torch.cuda.is_available(): model = model.to(DEFAULT_DEVICE)
model = model.cuda() video = video.to(DEFAULT_DEVICE)
video = video.cuda()
else:
print("CUDA is not available!")
pred_tracks, pred_visibility = model( pred_tracks, pred_visibility = model(
video, video,
@ -73,4 +73,4 @@ if __name__ == "__main__":
# save a video with predicted tracks # save a video with predicted tracks
seq_name = args.video_path.split("/")[-1] seq_name = args.video_path.split("/")[-1]
vis = Visualizer(save_dir="./saved_videos", pad_value=120, linewidth=3) vis = Visualizer(save_dir="./saved_videos", pad_value=120, linewidth=3)
vis.visualize(video, pred_tracks, query_frame=args.grid_query_frame) vis.visualize(video, pred_tracks, pred_visibility, query_frame=args.grid_query_frame)