Skip to content
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

Tobias/bg images and expr codes download #19

Open
wants to merge 2 commits into
base: main
Choose a base branch
from
Open

Tobias/bg images and expr codes download #19

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
c0059f0
Add background image download links + description
tobias-kirschstein Oct 1, 2024
cd6a495
Add expression code download links + description
tobias-kirschstein Oct 1, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 7 additions & 0 deletions DATASHEET.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -74,6 +74,10 @@ Datasheet in the format of "Datasheets for datasets" as described in
│ │   ├── cam{cam_name_02}.zip
│ │   ├── ...
│ │   └── cam{cam_name_80}.zip
│ ├── bg_image
│ │   └── bg_image.zip
│ ├── expression_codes
│ │   └── aeparams_1440000.pkl
│ ├── keypoints_3d
│ │   └── keypoints_3d.zip
│ ├── kinematic_tracking
Expand Down Expand Up @@ -109,6 +113,9 @@ Datasheet in the format of "Datasheets for datasets" as described in
In our releases, `N_CAMERAS` is 80, and `N_FRAMES` is either around 5,000 or 10,000.
The images follow the naming convention `{frame_id:06d}.avif`, as indicated in `frame_list.csv`, and are stored in avif format.
The images are RGB, with resolution 1024 x 667 or 2048 x 1334 depending on the release package.
* `decoder/bg_image/bg_image.zip` is a zip file with `N_CAMERAS` images that contain the empty background (without person) for each viewpoint which is helpful for background removal.
Background images are stored as `{cam_name}.avif`.
* `decoder/expression_codes/aeparams_1440000.pkl` is a pickled dictionary containing a mapping `frame_id => expression_code` where `expression_code` is a 256-dimensional code describing the person's facial expression at `frame_id`.
* `decoder/keypoints_3d/keypoints_3d.zip` is a zip file with `N_FRAMES` files. Each file follows the format `{frame_id:06d}.npy`, and is a npy file with a matrix with up to 274 rows 6 rows representing 3d keypoints for each frame.
Note that some keypoints may not be present if the number of inliers falls below a certain thresold. The format of each row is:
```
Expand Down
2 changes: 2 additions & 0 deletions download.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -34,6 +34,8 @@
"frame_list": ["frame_list.csv"],
"head_pose": ["head_pose/head_pose.zip"],
"image": ["image/cam{camera}.zip"],
"bg_image": ["bg_image/bg_image.zip"],
"expression_codes": ["expression_codes/aeparams_1440000.pkl"],
"keypoints_3d": ["keypoints_3d/keypoints_3d.zip"],
"kinematic_tracking": [
"kinematic_tracking/registration_vertices_mean.npy",
Expand Down
Loading