Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -25,8 +25,6 @@ Full-size realistic masks with pronounced “skin” texture. Sometimes they hav
|
|
| 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 |
|
| 28 |
-
|
| 29 |
-
|
| 30 |
# Why in-the-wild collection improves robustness
|
| 31 |
Crowd-sourced, in-the-wild data covers the natural long-tail diversity of real scenarios, so models trained on it generalize better and rely less on hidden shortcuts typical of controlled, in-office datasets
|
| 32 |
|
|
@@ -41,7 +39,7 @@ In in-house collected datasets, even with artificial variation of backgrounds/an
|
|
| 41 |
|
| 42 |
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
|
| 43 |
|
| 44 |
-
## Full version of dataset is availible for commercial usage - leave a request on our website [Axonlabs ](https://axonlabs.pro/?utm_source=
|
| 45 |
|
| 46 |
# Best Uses
|
| 47 |
This dataset is ideal for entities striving to meet or exceed iBeta Level 2 certification. By integrating this dataset, organizations can greatly enhance the training effectiveness of anti-spoofing algorithms, ensuring a robust and accurate performance in practical scenarios
|
|
|
|
| 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 |
|
|
|
|
|
|
|
| 28 |
# Why in-the-wild collection improves robustness
|
| 29 |
Crowd-sourced, in-the-wild data covers the natural long-tail diversity of real scenarios, so models trained on it generalize better and rely less on hidden shortcuts typical of controlled, in-office datasets
|
| 30 |
|
|
|
|
| 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
|
| 41 |
|
| 42 |
+
## Full version of dataset is availible for commercial usage - leave a request on our website [Axonlabs ](https://axonlabs.pro/?utm_source=hugging-face&utm_medium=cpc&utm_campaign=profile&utm_content=profile_link)to purchase the dataset 💰
|
| 43 |
|
| 44 |
# Best Uses
|
| 45 |
This dataset is ideal for entities striving to meet or exceed iBeta Level 2 certification. By integrating this dataset, organizations can greatly enhance the training effectiveness of anti-spoofing algorithms, ensuring a robust and accurate performance in practical scenarios
|