| ---
|
| license: apache-2.0
|
| tags: [point-tracking, video]
|
| ---
|
|
|
| # TinyPine Tracking
|
|
|
| The point-tracking model TinyPine Studio uses to follow points and areas through a video, unmodified,
|
| mirrored so the app's installer does not depend on a third-party download link. TinyPine did not train
|
| or change it.
|
|
|
| | Path | Model | Author | License |
|
| |---|---|---|---|
|
| | `bootstapir/causal_bootstapir_checkpoint.pt` | BootsTAPIR (online), PyTorch | Google DeepMind, [TAPNet](https://github.com/google-deepmind/tapnet) | Apache-2.0 (`bootstapir/LICENSE`) |
|
|
|
| The TAPNet README ("License and Disclaimer") states that all pre-trained checkpoints released in the
|
| repository, including BootsTAPIR, are licensed under Apache 2.0.
|
| Source: https://storage.googleapis.com/dm-tapnet/bootstap/causal_bootstapir_checkpoint.pt
|
|
|
| SHA-256: `bootstapir/causal_bootstapir_checkpoint.pt` 87c1e752cf5ce56e3e2f7da460aeb4d40fc826d04ef2939bade86a5c7495377f
|
|
|
