Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,50 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: cc-by-nc-4.0
|
| 3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-nc-4.0
|
| 3 |
+
---
|
| 4 |
+
# Anti-Spoofing dataset: live, replay, cut, print, 3D masks - large-scale face anti spoofing
|
| 5 |
+
This dataset delivers a single, end-to-end resource for training and benchmarking facial liveness-detection systems. By aggregating live sessions and eleven realistic presentation-attack classes into one collection, it accelerates development toward iBeta Level 1/2 compliance and strengthens model robustness against the full spectrum of spoofing tactics
|
| 6 |
+
|
| 7 |
+
|
| 8 |
+
|
| 9 |
+
##Why Comprehensive Anti-Spoofing Data?
|
| 10 |
+
Modern certification pipelines demand proof that a system resists all common attack vectors—not just prints or replays. This dataset delivers those vectors in one place, allowing you to:
|
| 11 |
+
- Benchmark a model’s true generalisation
|
| 12 |
+
- Fine-tune against rare but high-impact threats (e.g., silicone or textile masks)
|
| 13 |
+
- Streamline audits by demonstrating coverage of every ISO 30107-3 attack category
|
| 14 |
+
|
| 15 |
+
##Dataset Features
|
| 16 |
+
- **Dataset Size:** ≈ 95 000 videos / image sequences spanning live captures and eleven spoof classes
|
| 17 |
+
- **Attack Diversity:** 3D paper mask, wrapped 3D mask, photo print, mobile replay, display replay, cut-out 2D mask, silicone mask, latex mask, textile mask
|
| 18 |
+
- **Active Liveness Cues:** Natural blinks, and head rotations included across live and mask sessions
|
| 19 |
+
- **Attribute Range:** different combinations of hairstyles, eyewear, facial hair, and accessories.
|
| 20 |
+
- **Environmental Variability:** Indoor/outdoor scenes under various lighting conditions
|
| 21 |
+
- **Multi-angle Capture:** Mainly used selfie camera, also back
|
| 22 |
+
- **Capture Devices:** Footage from flagship and mid-range phones (iPhone 14 / 13 Pro, Galaxy S23, Pixel 7, Redmi Note 12 Pro+, Galaxy A54, Honor 70)
|
| 23 |
+
- **Additional Flexibility:** Custom re-captures available on request
|
| 24 |
+
|
| 25 |
+
## Full version of dataset is availible for commercial usage - leave a request on our website [Axonlabs ](https://axonlabs.pro/?utm_source=huggingface&utm_medium=dataset_card&utm_campaign=display_replay_attacks&utm_id=12345)to purchase the dataset 💰
|
| 26 |
+
|
| 27 |
+
## Technical Specifications
|
| 28 |
+
- **File Format:** MP4 for video, JPEG/PNG for still sequences; all compatible with mainstream ML frameworks
|
| 29 |
+
- **Resolution & FPS:** Up to 4K @ 60 fps; balanced presets included for rapid training
|
| 30 |
+
|
| 31 |
+
## Best Uses
|
| 32 |
+
Ideal for companies pursuing or maintaining iBeta Level 1/2 certification, research groups exploring new PAD architectures, and vendors stress-testing production face-verification pipelines
|
| 33 |
+
|
| 34 |
+
## Attack Classes
|
| 35 |
+
- [Live / Genuine](https://huggingface.co/datasets/AxonData/Selfie_and_Official_ID_Photo_Dataset) Natural faces with spontaneous movements across varied devices and lighting
|
| 36 |
+
- [3D Paper Mask](https://huggingface.co/datasets/AxonData/3D_paper_mask_attack_dataset_for_Liveness) Folded paper masks with protruding nose/forehead
|
| 37 |
+
- [Wrapped 3D Print](https://huggingface.co/datasets/AxonData/Wrapped_3D_Attacks) Rigid paper moulds reproducing head geometry
|
| 38 |
+
- [Photo Print](https://www.kaggle.com/datasets/axondata/photo-print-attacks-dataset-1k-individuals) Glossy still photos at multiple angles—the classic 2D spoof
|
| 39 |
+
- Cylinder 3D Paper Mask A folded or cylindrical sheet of paper that simulates volume
|
| 40 |
+
- [Mobile Replay](https://huggingface.co/datasets/AxonData/Replay_attack_mobile) Face videos played on phone screens; includes glare and auto-brightness shifts
|
| 41 |
+
- [Display Replay](https://huggingface.co/datasets/AxonData/Display_replay_attacks) Attacks via monitors, and laptops
|
| 42 |
+
- [Cut-out 2D Mask](https://huggingface.co/datasets/AxonData/Anti_Spoofing_Cut_print_attack) Flat printed masks with eye/mouth holes plus active head motion
|
| 43 |
+
- On-actor Print / Cuts Paper elements (photos, cutouts) are glued directly onto the actor's face
|
| 44 |
+
- [Silicone](https://huggingface.co/datasets/AxonData/iBeta_level_2_Silicone_masks) and [Latex Masks](https://huggingface.co/datasets/AxonData/Latex_Mask_dataset) High-detail silicone/latex overlays with blinking and subtle mimicry
|
| 45 |
+
- [Cloth 3D Mask](https://huggingface.co/datasets/AxonData/3d_cloth_face_mask_spoofing_dataset) Elastic fabric masks hugging facial contours during movement
|
| 46 |
+
- [High-Fidelity Resin Mask](https://huggingface.co/datasets/AxonData/high_precision_3d_resin_masks_real_faces Hyperrealistic masks with detailed skin texture
|
| 47 |
+
|
| 48 |
+
|
| 49 |
+
## Conclusion
|
| 50 |
+
This dataset’s scale, breadth of attack types, and real-world capture conditions make it indispensable for anyone building or evaluating biometric anti-spoofing solutions. Deploy it to harden your systems against today’s—and tomorrow’s—most sophisticated presentation attacks
|