Silicon23/itw_pipeline_visualization

ITW Pipeline Visualization — assets

Purpose of this repository

This repository exists for one reason: to serve media files to the visualization site at https://silicon23.github.io/itw_pipeline_visualization/. A static site cannot host its own heavy media, so the frames, depth maps and overlay renders it streams live here.

It is not published for redistribution, and it is not a dataset to train or evaluate on. It is the asset backing of a figure — the equivalent of the images embedded in a paper, stored where a browser can fetch them.

Ego4D-derived assets are shown under the Ego4D licence's publication-website provision, which permits reproducing images and videos from the Database in a research or academic publication "or any website through which such publication(s) is made available". They may not be redistributed, and may not be used as a dataset. Ego4D remains the property of its licensor; anyone wanting the underlying data must obtain their own signed Ego4D licence.

Rendered assets for the visualization site of an in-the-wild open-vocabulary 3D video object tracking project. Each pipeline takes an RGB video plus a text category label and produces a per-frame 3D bounding-box trajectory; the site shows what each one made of the same clip.

This repository holds only the heavy media the site streams at runtime: per-frame depth and colour for the browser-side point cloud, overlay renders, and gallery previews. The site itself is static and hosted separately; it fetches these files cross-origin.

These are derived visualization artifacts, not training data. Every frame originates from a source dataset — see Datasets and attribution.

Datasets and attribution

Assets are grouped by source dataset. Each dataset's assets carry that dataset's own licence and attribution, and the most restrictive terms govern their use. This table is the authoritative list of what is present.

id	source dataset	licence	use permitted here
sav	SA-V (Segment Anything Video, Meta AI)	CC BY 4.0	redistribution permitted with attribution — Ravi et al., SAM 2: Segment Anything in Images and Videos, 2024
sasafari	SA-FARI (Meta AI & Conservation X Labs)	CC BY-NC 4.0	redistribution permitted with attribution, non-commercial use only — The SA-FARI Dataset, arXiv:2511.15622
ego4d	Ego4D EgoTracks	signed-licence dataset	serve-to-site only under the publication-website provision; no redistribution, not a dataset

Ego4D attribution, as its licence requires: "Egocentric Live 4D Perception (Ego4D) Database: A large-scale first-person video database, supporting research in multi-modal machine perception for daily life activity. Ego4D Consortium 2020. https://sites.google.com/view/ego4d/home".

The repository-level licence tag is other because the sources carry different terms. Read the table above rather than the tag.

Layout

assets/<dataset_id>/<video_id>/
├── preview.mp4 · poster.jpg · strip.jpg        gallery preview
└── <pipeline_id>/
    ├── manifest.json          frames, intrinsics, metric camera-to-world poses
    ├── boxes.json             per-object per-frame 3D boxes (see conventions)
    ├── visualizations.json    what overlays exist, and at what cadence
    ├── depth/000000.webp …    uint16 metric depth, packed into R and G
    ├── rgb/000000.jpg …       colour for the point cloud
    └── overlays/<name>/
        ├── video.mp4          for playback
        └── frames/NNNNNN.jpg  stills at that overlay's own cadence

Pipelines currently present: wilddet3d (prompt-and-detect). Overlays currently present: input (source video with ground-truth masks), box_prompt, point_prompt.

Conventions

These hold across every dataset and pipeline in this repository.

Depth. WebP with the 16-bit value split across the red and green channels, because no single-channel 16-bit format decodes reliably across browsers:

z_metres = (R * 256 + G) / 65535 * depth_max_m      # z == 0 means invalid

depth_max_m is per clip, in manifest.json. Depth is metric metres, camera-Z, OpenCV axes (x right, y down, z forward).

Boxes. Camera frame — the detector applies no extrinsic. Multiply by the per-frame c2w in manifest.json for world frame. Each box carries center, dims_xyz (extents along the box's own axes), quat_wxyz, and corners.

Corner order is canonical and identical across pipelines:

0 (-,-,-)  1 (+,-,-)  2 (+,+,-)  3 (-,+,-)
4 (-,-,+)  5 (+,-,+)  6 (+,+,+)  7 (-,+,+)

Edges: bottom 0-1-2-3-0, top 4-5-6-7-4, verticals 0-4 1-5 2-6 3-7. Corners are validated elementwise against center + (signs · dims/2) @ Rᵀ before export, because IoU is order-invariant and cannot catch an ordering mistake.

Frame indices are always native source-video frames, so overlays at different cadences share one timeline. Every overlay video is encoded at the source frame rate with its content held between updates, so all overlays for a clip have identical duration and stay in sync under a single playback rate.

Annotation cadence is preserved, not interpolated. Where a source dataset annotates below video rate — SA-V labels every 4th frame at 24 fps, giving 6 Hz masks — the overlay holds the previous annotation rather than inventing intermediate ground truth, and visualizations.json records the true cadence.