Notation and Symbols for Deepfake Detection Algorithm

A 5.5 KB Excel file detailing the notation and symbols used in the ARC-Net deepfake detection algorithm. The dataset was authored by Md Shihab Reza and last updated in April 2026. It supports a study that trained models on a combined dataset of 500 real images from Bangladesh and 140,000 real and fake faces.

Use Cases

• Map algorithm components like attention and residual convolutional layers to their formal mathematical symbols for model documentation.
• Interpret statistical test results (e.g., McNemar's test) reported in the associated research paper using the defined notation.
• Replicate the ARC-Net model architecture or ablation studies by understanding the formal symbols for its building blocks.
• Cross-reference Explainable AI (XAI) technique outputs (Grad-CAM, LIME) with the algorithm's symbolic framework for analysis.

Strengths

• Clear provenance linked to a published research paper achieving 99% accuracy in deepfake detection.
• Released under a permissive CC-BY-4.0 license for open use and modification.

Limitations

• Extremely small scale at 5.5 KB, containing only symbolic definitions without the underlying image or model data.
• No sample data or column definitions are provided, limiting immediate analytical utility.

Provenance

Source

Md Shihab Reza, associated with the ARC-Net deepfake detection research.

Collection Method

Created as supplementary material defining the notation for a published algorithm.

Freshness

Last updated in April 2026.

Geography

Study context includes facial image data from Bangladesh.