ljchang commited on
Commit
bd515c9
·
verified ·
1 Parent(s): a97b480

Clarify model licensing scope and dataset provenance

Browse files

Separate implementation licenses from pretrained-weight and training-data terms. Preserve existing valid grants and document unresolved distribution permissions. No weights or access settings changed.

Files changed (2) hide show
  1. LICENSE-NOTICE.md +7 -0
  2. README.md +13 -2
LICENSE-NOTICE.md ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ # img2pose — licensing scope notice
2
+
3
+ The upstream img2pose code and model distribution use **CC BY-NC 4.0**; see the [img2pose upstream license](https://github.com/vitoralbiero/img2pose/blob/main/license.md). This repository's existing `license: cc-by-nc-4.0` metadata is retained. Preserve attribution, the license reference, and notices of modifications or format conversion. Py-Feat's package-level MIT license does not override these terms.
4
+
5
+ The [upstream project](https://github.com/vitoralbiero/img2pose) documents WIDER FACE training and separately fine-tuned checkpoints. The original-checkpoint variant and training-data permissions must be identified for each converted artifact. CC BY-NC 4.0 is a noncommercial license, not an unrestricted permissive license.
6
+
7
+ This notice clarifies scope; it does not revoke existing valid grants or create a new license for third-party material. Dataset and teacher terms do not automatically relicense every trained artifact or inference output. A software license or model-card badge alone does not establish all checkpoint redistribution or commercial-use permissions.
README.md CHANGED
@@ -9,6 +9,8 @@ license: cc-by-nc-4.0
9
 
10
  # img2pose
11
 
 
 
12
  ## Model Description
13
  img2pose uses Faster R-CNN to predict 6 Degree of Freedom Pose (DoF) for all faces in the photo. An interesting property of this model is that it can project the 3D face onto a 2D plane to also identify bounding boxes for each face. It does not require any other face detection model.
14
 
@@ -37,7 +39,7 @@ Vítor Albiero, Xingyu Chen, Xi Yin, Guan Pang, Tal Hassner, "img2pose: Face Ali
37
  ```
38
 
39
  ## Acknowledgements
40
- We thank Albiero Vítor for sharing their code and training weights with a permissive license.
41
 
42
  ## Example Useage
43
 
@@ -93,4 +95,13 @@ bbox = img2pose_output['boxes']
93
  poses = img2pose_output['dofs']
94
  facescores = img2pose_output['scores']
95
 
96
- ```
 
 
 
 
 
 
 
 
 
 
9
 
10
  # img2pose
11
 
12
+ > **Licensing scope:** See the [license and provenance notice](LICENSE-NOTICE.md) before relying on this card's license metadata for pretrained-weight redistribution or commercial use. Existing valid grants are preserved.
13
+
14
  ## Model Description
15
  img2pose uses Faster R-CNN to predict 6 Degree of Freedom Pose (DoF) for all faces in the photo. An interesting property of this model is that it can project the 3D face onto a 2D plane to also identify bounding boxes for each face. It does not require any other face detection model.
16
 
 
39
  ```
40
 
41
  ## Acknowledgements
42
+ We thank Vítor Albiero and colleagues for sharing their code and trained weights under the upstream CC BY-NC 4.0 terms.
43
 
44
  ## Example Useage
45
 
 
95
  poses = img2pose_output['dofs']
96
  facescores = img2pose_output['scores']
97
 
98
+ ```
99
+
100
+ ## License and provenance
101
+
102
+ The upstream img2pose code and model distribution use **CC BY-NC 4.0**; see the [img2pose upstream license](https://github.com/vitoralbiero/img2pose/blob/main/license.md). This repository's existing `license: cc-by-nc-4.0` metadata is retained. Preserve attribution, the license reference, and notices of modifications or format conversion. Py-Feat's package-level MIT license does not override these terms.
103
+
104
+ The [upstream project](https://github.com/vitoralbiero/img2pose) documents WIDER FACE training and separately fine-tuned checkpoints. The original-checkpoint variant and training-data permissions must be identified for each converted artifact. CC BY-NC 4.0 is a noncommercial license, not an unrestricted permissive license.
105
+
106
+ This notice clarifies scope; it does not revoke existing valid grants or create a new license for third-party material. Dataset and teacher terms do not automatically relicense every trained artifact or inference output. A software license or model-card badge alone does not establish all checkpoint redistribution or commercial-use permissions.
107
+