Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -22,6 +22,8 @@ Nylon/elastic fabric mask; full-color face print, fabric fits snugly around the
|
|
| 22 |
- # Latex — volume 3D latex masks
|
| 23 |
Full-size realistic masks with pronounced “skin” texture. Sometimes they have cut-out eyes and are complemented by external attributes
|
| 24 |
|
|
|
|
|
|
|
| 25 |
# Why is this combination of attacks optimal?
|
| 26 |
They cover three different types of attacks (flat printing → textile 3D deformation → realistic 3D object), which allows testing the resistance of models to increasing attack realism
|
| 27 |
|
|
@@ -35,6 +37,8 @@ Crowd-sourced, in-the-wild data covers the natural long-tail diversity of real s
|
|
| 35 |
- Broad spectrum of PAIs/masks. Different materials, shapes, and application methods; real-world artifacts (glare, folds, misalignment); plus “imperfect” spoofing attempts
|
| 36 |
- Demographic and cultural diversity. Skin tones, makeup, accessories (glasses, headwear), styles—rarely covered in office setups
|
| 37 |
|
|
|
|
|
|
|
| 38 |
In in-house collected datasets, even with artificial variation of backgrounds/angles, common constants remain: the same camera pool, typical lighting and backgrounds, repeated rooms and collection crews. Models quickly latch onto these spurious cues (e.g., characteristic white balance or wall texture), which harms transfer to real-world conditions
|
| 39 |
|
| 40 |
Because the sources of variability differ, in-the-wild and in-house datasets have weak overlap. This makes our dataset a valuable external test bed: if a model performs well here, the likelihood of failure in real production is substantially lower
|
|
|
|
| 22 |
- # Latex — volume 3D latex masks
|
| 23 |
Full-size realistic masks with pronounced “skin” texture. Sometimes they have cut-out eyes and are complemented by external attributes
|
| 24 |
|
| 25 |
+
|
| 26 |
+
|
| 27 |
# Why is this combination of attacks optimal?
|
| 28 |
They cover three different types of attacks (flat printing → textile 3D deformation → realistic 3D object), which allows testing the resistance of models to increasing attack realism
|
| 29 |
|
|
|
|
| 37 |
- Broad spectrum of PAIs/masks. Different materials, shapes, and application methods; real-world artifacts (glare, folds, misalignment); plus “imperfect” spoofing attempts
|
| 38 |
- Demographic and cultural diversity. Skin tones, makeup, accessories (glasses, headwear), styles—rarely covered in office setups
|
| 39 |
|
| 40 |
+
|
| 41 |
+
|
| 42 |
In in-house collected datasets, even with artificial variation of backgrounds/angles, common constants remain: the same camera pool, typical lighting and backgrounds, repeated rooms and collection crews. Models quickly latch onto these spurious cues (e.g., characteristic white balance or wall texture), which harms transfer to real-world conditions
|
| 43 |
|
| 44 |
Because the sources of variability differ, in-the-wild and in-house datasets have weak overlap. This makes our dataset a valuable external test bed: if a model performs well here, the likelihood of failure in real production is substantially lower
|