Downloads · 30 days
6
5% of all-time downloads
deepfake-deit
ChristianDW15/deepfake-deit
deepfake-deit is a image classification model from ChristianDW15. Use it when you need a label for an image. It is set up for transformers.
Model DeiT untuk deteksi citra deepfake. Versi ini adalah Stage 3 DCT: patch embedding dimodifikasi Conv2d(3→4) dengan channel ke-4 = fitur DCT (frekuensi). Bukan model RGB standar — input harus 4-channel (RGB + DCT).
All-time downloads
128
Public
Parameters
86M
1 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors344 MB · 100%
At a glance
- Task
- Image Classification
- Library
- transformers
- Model type
- deit
- Access
- Public
- Created
- May 17, 2026
- Updated
- Jul 1, 2026
- SHA
- a7e4cf96
Model card
From the Hugging Face model README
Deepfake Detection DeiT (Stage 3, DCT 4-channel)
Model DeiT untuk deteksi citra deepfake. Versi ini adalah Stage 3 DCT:
patch embedding dimodifikasi Conv2d(3→4) dengan channel ke-4 = fitur DCT
(frekuensi). Bukan model RGB standar — input harus 4-channel (RGB + DCT).
Label
0 = FAKE, 1 = REAL (ImageFolder alfabetis: Fake/ < Real/).
Threshold
Gunakan 0.45177 (Youden's J) pada P(Real) — bukan 0.5.
Klasifikasi REAL bila P(Real) >= 0.45177, selain itu FAKE.
Cara load (custom, BUKAN from_pretrained)
from_pretrained() standar tidak cocok karena patch embed 4-channel.
Load via kode kustom: bangun arsitektur DCT (patch embed 3→4) lalu
load_state_dict(safetensors.torch.load_file("model.safetensors")).
Preprocessing wajib menambah channel DCT (cv2.dct → log1p(abs) → minmax)
identik dengan pipeline training. Widget inferensi HF dinonaktifkan
(inference: false) karena butuh preprocessing 4-channel kustom.
- Task
- Image Classification
- Library
- transformers
- Type
- deit
- Created
- May 17, 2026
- Updated
- Jul 1, 2026