Cdf video data augmentation vs Physical AI video data augmentation
Cdf video data augmentation is an open-source agent skill for design and media. Physical AI video data augmentation is an open-source agent skill for design and media.
- Cdf video data augmentation is best for generate weather and time-of-day variants of my driving videos with cosmos transfer.
- Physical AI video data augmentation is best for augment a set of driving videos and auto-label the augmented outputs.
- What it is
- Cdf video data augmentation is an open-source agent skill for design and media.Physical AI video data augmentation is an open-source agent skill for design and media.
- Best for
- Generate weather and time-of-day variants of my driving videos with Cosmos TransferAugment a set of driving videos and auto-label the augmented outputs
- Kind
- SkillSkill
- Pricing
- Open sourceOpen source
- Platforms
- Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, Windsurf, GooseClaude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, Windsurf, Goose
- GitHub stars
- 4K4K
- Last commit
- Nov 2026Nov 2026
- License
- Apache-2.0Apache-2.0
- Open source
- Good for
- Generate weather and time-of-day variants of my driving videos with Cosmos Transfer; Check whether my Airflow cluster has the pools and GPUs needed for a video augmentation run; Trigger a bounded run on 10 source videos and monitor it until it finishesAugment a set of driving videos and auto-label the augmented outputs; Pseudo-label source videos without any augmentation step; Run the full labeling pipeline on a GPU pool and download the results
What Cdf video data augmentation does
Payload authoring and validation
Renders and checks the DagRun payload for Kubernetes or OSMO before any run is submitted.
Preflight readiness checks
Verifies DAG loading, pools, GPU availability, endpoints and storage mapping before triggering.
Run submission and monitoring
Triggers a single Airflow DagRun and tracks its task instances through to completion.
Output verification
Confirms expected candidate, labeling and training-export artifacts exist and decode correctly.
Video upload helper
Validates and optionally uploads a local video file or flat directory to object storage.
Result summary script
Summarizes a downloaded run's candidates, evaluator results and exports without loading media.
What Physical AI video data augmentation does
Workflow selection
Maps the user's request to one of four OSMO workflow variants for labeling, augmentation or both.
Preflight and readiness checks
Verifies secrets, registry image access, GPU pool status and model cache before submission.
Submit-time interpolation
Derives storage and cache paths from the active dataset backend and passes them in one submit command.
Run monitoring
Queries workflow status, task states and logs through the OSMO CLI until the run completes.
Output retrieval and comparison
Downloads run outputs and renders side-by-side comparisons of original and augmented videos.
Demo asset fallback
Uses the published VDA demo dataset when the user provides no input video.
Cdf video data augmentation vs Physical AI video data augmentation: questions
Is Cdf video data augmentation or Physical AI video data augmentation cheaper?
Both are open source. Compare current plans on each official site.
What is the difference between Cdf video data augmentation and Physical AI video data augmentation?
Cdf video data augmentation is an open-source agent skill for design and media. Physical AI video data augmentation is an open-source agent skill for design and media.
