Summary This repository serves as a comprehensive, curated list of academic research, tools, and code related to fake audio detection. It aims to provide researchers, engineers, and practitioners with an up-to-date overview of the rapidly evolving field of deepfake audio detection, enabling them to stay abreast of the latest advancements and identify relevant resources for their work.

How It Works The project functions as a dynamic bibliography, meticulously compiling and organizing research papers, primarily from arXiv, that address various facets of fake audio detection. It covers novel detection algorithms, forensic analysis, dataset creation, challenge evaluations, and defense mechanisms against sophisticated audio manipulation techniques. The collection's strength lies in its breadth and recency, offering a centralized point of reference for the latest scientific contributions and tracking research trends from early methods to advanced techniques leveraging self-supervised learning, LLMs, and multimodal approaches.

Quick Start & Requirements This repository is a curated list of research papers and does not contain executable code or software requiring installation. Users are directed to the linked arXiv papers for details on specific methodologies, datasets, and implementation requirements.

Highlighted Details

• Features an extensive collection of research papers, with entries dating from 2019 to May 2026, reflecting the field's rapid development and the continuous emergence of new techniques and challenges.
• Covers a broad range of fake audio detection sub-topics, including real-time detection, environmental sound deepfakes, forensic analysis, source tracing, and detection across various languages and domains, focusing on robustness, generalization, and neural audio codecs.
• Includes entries related to significant academic challenges and benchmarks like ASVspoof and ADD, indicating a focus on standardized evaluation and real-world applicability.
• Highlights research exploring advanced techniques such as self-supervised learning, contrastive learning, diffusion models, and the application of large language models (LLMs) for audio deepfake detection.

Maintenance & Community The repository shows recent activity, with updates noted as of May 6, 2026, suggesting active curation. The community appears to be primarily academic researchers and developers. No specific community links are provided.

Licensing & Compatibility No specific open-source license is mentioned for the repository itself. The linked research papers are typically available under open-access licenses or terms set by arXiv. Commercial use depends on the licensing of individual papers or associated code.

Limitations & Caveats: As a curated list, this repository does not provide ready-to-use tools or implementations; users must access and implement the research papers independently. The focus is heavily on academic publications, potentially omitting commercial solutions or practical deployment guides. The rapid pace of deepfake technology means the list requires continuous updates to remain fully comprehensive.