8000 Cannot move predicted nodes after training in SLEAP GUI (v1.4.1) · Issue #2154 · talmolab/sleap · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

Cannot move predicted nodes after training in SLEAP GUI (v1.4.1) #2154

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Open
1 task done
KeYamaoka opened this issue Apr 7, 2025 · 0 comments
Open
1 task done

Cannot move predicted nodes after training in SLEAP GUI (v1.4.1) #2154

KeYamaoka opened this issue Apr 7, 2025 · 0 comments
Labels
bug Something isn't working

Comments

@KeYamaoka
Copy link

Bug description

In SLEAP GUI (v1.4.1 on Linux), predicted nodes cannot be manually moved or adjusted by double-clicking, as expected. This prevents correction after inference.

Expected behaviour

Double-clicking on a predicted node should enable it to be moved for manual correction (just like user-labeled instances).

Actual behaviour

After training, SLEAP automatically generates predicted instances for the previously unlabeled frames in the training video. These predicted nodes are displayed in the GUI, but they cannot be edited or moved by double-clicking, as expected.

In addition, the menu item "Edit > Convert predicted to user-labeled" does not exist in the GUI, so there is no way to convert these predicted nodes into editable labels.

Your personal set up

  • OS: Linux (Linux-5.4.0-212-generic-x86_64-with-debian-bullseye-sid)

  • Python Version: 3.7.12

  • SLEAP Version: 1.4.1

  • TensorFlow Version: 2.8.4

  • Numpy Version: 1.21.6

  • Installation method:

    • pip package (inside virtual environment)

  • GPU: NVIDIA RTX A6000

  • Driver Version: 530.41.03

  • CUDA Version: 12.1

  • GPU Memory: 49140 MiB

Screenshots

How to reproduce

  1. Annotate 16 frames manually in a video (e.g., with 3 mice using centered_instance + centroid).
  2. Leave 2 frames unlabeled in the same video.
  3. Run training in SLEAP GUI.
  4. After training finishes, predicted nodes are automatically shown on the 2 unlabeled frames.
  5. Try to double-click and move the predicted nodes.
  6. Nothing happens — the nodes are not selectable or editable.
@KeYamaoka KeYamaoka added the bug Something isn't working label Apr 7, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
bug Something isn't working
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@KeYamaoka
0