EchoFake: A Replay-Aware Dataset for Speech Deepfake Detection

A dataset designed for detecting speech deepfakes in practical scenarios like telephone fraud and identity theft. It was created by EchoFake and last updated on March 2, 2026. The dataset includes baseline model code and auto-recording tools.

Use Cases

• Train models to detect replay attacks based on the dataset's focus on practical scenarios.
• Benchmark speech deepfake detection algorithms using the provided baseline models.
• Simulate telephone fraud detection systems based on the described real-world application context.
• Study identity theft prevention techniques using speech authentication data.

Strengths

• Includes baseline model code and auto-recording tools, facilitating reproducibility and experimentation.
• Focuses on practical, real-world attack scenarios like telephone fraud, increasing applied relevance.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Last updated 2026-03-02 02:42:50; freshness should be verified.

Provenance

Source

EchoFake

Freshness

Last updated 2026-03-02 02:42:50.