Mengieong/SEED_balanced
SEED_balanced SEED_balanced is the public balanced release of SEED, a benchmark for provenance tracing in sequential deepfake facial edits. Unlike conventional deepfake datasets that focus on single-step manipulations or binary real/fake detection, SEED models multi-step diffusion-based facial editing trajectories and supports three complementary tasks: Authenticity Analysis, Editing Trace Analysis, and Spatial Evidence Analysis. The full SEED benchmark contains 91,526 images… See the full description on the dataset page: https://huggingface.co/datasets/Mengieong/SEED_balanced.
Add self-contained viewer parquet files and README viewer config
Add viewer preview Parquet files and README viewer config
Add viewer preview Parquet files and README viewer config
Update README.md
Upload folder using huggingface_hub
Upload SEED_subset_1.zip with huggingface_hub
Upload SEED_subset_3.zip with huggingface_hub
Update README.md
Update README.md
Upload prediction.zip
Delete dim512_finetuned_test_predictions.csv
Delete SEED_dataset.zip
Upload predictions template CSV
Upload SEED dataset zip
Delete seqdeepfake_train_data.zip
Update README.md
Update README.md
Upload seqdeepfake_train_data.zip with huggingface_hub
Upload 253 files
Delete seqdeepfake_train_data.zip
Update README.md
Update README.md
Upload 2 files
Create assets/pd.md
Delete assets
Create assets/
Create README.md
Upload seqdeepfake_train_data.zip with huggingface_hub
initial commit