Audio generation systems now create very realistic soundscapes that can enhance media production, but also pose potential risks. Several studies have examined deepfakes in speech or singing voice. However, environmental sounds have different characteristics, which may make methods for detecting speech and singing deepfakes less effective for real-world sounds. In addition, existing datasets for environmental sound deepfake detection are limited in scale and audio types. To address this gap, we introduce EnvSDD, the first large-scale curated dataset designed for this task, consisting of 45.25 hours of real and 316.74 hours of fake audio. The test set includes diverse conditions to evaluate the generalizability, such as unseen generation models and unseen datasets. We also propose an audio deepfake detection system, based on a pre-trained audio foundation model. Results on EnvSDD show that our proposed system outperforms the state-of-the-art systems from speech and singing domains.
Recent advances in text-to-speech technologies have enabled realistic voice generation, fueling audio-based deepfake attacks such as fraud and impersonation. While audio anti-spoofing systems are critical for detecting such threats, prior work has predominantly focused on acoustic-level perturbations, leaving the impact of linguistic variation largely unexplored. In this paper, we investigate the linguistic sensitivity of both open-source and commercial anti-spoofing detectors by introducing transcript-level adversarial attacks. Our extensive evaluation reveals that even minor linguistic perturbations can significantly degrade detection accuracy: attack success rates surpass 60% on several open-source detector-voice pairs, and notably one commercial detection accuracy drops from 100% on synthetic audio to just 32%. Through a comprehensive feature attribution analysis, we identify that both linguistic complexity and model-level audio embedding similarity contribute strongly to detector vulnerability. We further demonstrate the real-world risk via a case study replicating the Brad Pitt audio deepfake scam, using transcript adversarial attacks to completely bypass commercial detectors. These results highlight the need to move beyond purely acoustic defenses and account for linguistic variation in the design of robust anti-spoofing systems. All source code will be publicly available.
A key research area in deepfake speech detection is source tracing - determining the origin of synthesised utterances. The approaches may involve identifying the acoustic model (AM), vocoder model (VM), or other generation-specific parameters. However, progress is limited by the lack of a dedicated, systematically curated dataset. To address this, we introduce STOPA, a systematically varied and metadata-rich dataset for deepfake speech source tracing, covering 8 AMs, 6 VMs, and diverse parameter settings across 700k samples from 13 distinct synthesisers. Unlike existing datasets, which often feature limited variation or sparse metadata, STOPA provides a systematically controlled framework covering a broader range of generative factors, such as the choice of the vocoder model, acoustic model, or pretrained weights, ensuring higher attribution reliability. This control improves attribution accuracy, aiding forensic analysis, deepfake detection, and generative model transparency.
Deepfake (DF) detectors face significant challenges when deployed in real-world environments, particularly when encountering test samples deviated from training data through either postprocessing manipulations or distribution shifts. We demonstrate postprocessing techniques can completely obscure generation artifacts presented in DF samples, leading to performance degradation of DF detectors. To address these challenges, we propose Think Twice before Adaptation (\texttt{T$^2$A}), a novel online test-time adaptation method that enhances the adaptability of detectors during inference without requiring access to source training data or labels. Our key idea is to enable the model to explore alternative options through an Uncertainty-aware Negative Learning objective rather than solely relying on its initial predictions as commonly seen in entropy minimization (EM)-based approaches. We also introduce an Uncertain Sample Prioritization strategy and Gradients Masking technique to improve the adaptation by focusing on important samples and model parameters. Our theoretical analysis demonstrates that the proposed negative learning objective exhibits complementary behavior to EM, facilitating better adaptation capability. Empirically, our method achieves state-of-the-art results compared to existing test-time adaptation (TTA) approaches and significantly enhances the resilience and generalization of DF detectors during inference. Code is available \href{https://github.com/HongHanh2104/T2A-Think-Twice-Before-Adaptation}{here}.
In recent years, the rapid development of deepfake technology has given rise to an emerging and serious threat to public security: diffusion model-based digital human generation. Unlike traditional face manipulation methods, such models can generate highly realistic videos with consistency through multimodal control signals. Their flexibility and covertness pose severe challenges to existing detection strategies. To bridge this gap, we introduce DigiFakeAV, the first large-scale multimodal digital human forgery dataset based on diffusion models. Employing five latest digital human generation methods (Sonic, Hallo, etc.) and voice cloning method, we systematically produce a dataset comprising 60,000 videos (8.4 million frames), covering multiple nationalities, skin tones, genders, and real-world scenarios, significantly enhancing data diversity and realism. User studies show that the confusion rate between forged and real videos reaches 68%, and existing state-of-the-art (SOTA) detection models exhibit large drops in AUC values on DigiFakeAV, highlighting the challenge of the dataset. To address this problem, we further propose DigiShield, a detection baseline based on spatiotemporal and cross-modal fusion. By jointly modeling the 3D spatiotemporal features of videos and the semantic-acoustic features of audio, DigiShield achieves SOTA performance on both the DigiFakeAV and DF-TIMIT datasets. Experiments show that this method effectively identifies covert artifacts through fine-grained analysis of the temporal evolution of facial features in synthetic videos.
The proliferation of diffusion-based deepfake technologies poses significant risks for unauthorized and unethical facial image manipulation. While traditional countermeasures have primarily focused on passive detection methods, this paper introduces a novel proactive defense strategy through adversarial attacks that preemptively protect facial images from being exploited by diffusion-based deepfake systems. Existing adversarial protection methods predominantly target conventional generative architectures (GANs, AEs, VAEs) and fail to address the unique challenges presented by diffusion models, which have become the predominant framework for high-quality facial deepfakes. Current diffusion-specific adversarial approaches are limited by their reliance on specific model architectures and weights, rendering them ineffective against the diverse landscape of diffusion-based deepfake implementations. Additionally, they typically employ global perturbation strategies that inadequately address the region-specific nature of facial manipulation in deepfakes.
The rapid emergence of multimodal deepfakes (visual and auditory content are manipulated in concert) undermines the reliability of existing detectors that rely solely on modality-specific artifacts or cross-modal inconsistencies. In this work, we first demonstrate that modality-specific forensic traces (e.g., face-swap artifacts or spectral distortions) and modality-shared semantic misalignments (e.g., lip-speech asynchrony) offer complementary evidence, and that neglecting either aspect limits detection performance. Existing approaches either naively fuse modality-specific features without reconciling their conflicting characteristics or focus predominantly on semantic misalignment at the expense of modality-specific fine-grained artifact cues. To address these shortcomings, we propose a general multimodal framework for video deepfake detection via Cross-Modal Alignment and Distillation (CAD). CAD comprises two core components: 1) Cross-modal alignment that identifies inconsistencies in high-level semantic synchronization (e.g., lip-speech mismatches); 2) Cross-modal distillation that mitigates feature conflicts during fusion while preserving modality-specific forensic traces (e.g., spectral distortions in synthetic audio). Extensive experiments on both multimodal and unimodal (e.g., image-only/video-only)deepfake benchmarks demonstrate that CAD significantly outperforms previous methods, validating the necessity of harmonious integration of multimodal complementary information.
We show how replay attacks undermine audio deepfake detection: By playing and re-recording deepfake audio through various speakers and microphones, we make spoofed samples appear authentic to the detection model. To study this phenomenon in more detail, we introduce ReplayDF, a dataset of recordings derived from M-AILABS and MLAAD, featuring 109 speaker-microphone combinations across six languages and four TTS models. It includes diverse acoustic conditions, some highly challenging for detection. Our analysis of six open-source detection models across five datasets reveals significant vulnerability, with the top-performing W2V2-AASIST model's Equal Error Rate (EER) surging from 4.7% to 18.2%. Even with adaptive Room Impulse Response (RIR) retraining, performance remains compromised with an 11.0% EER. We release ReplayDF for non-commercial research use.
As deepfake speech becomes common and hard to detect, it is vital to trace its source. Recent work on audio deepfake source tracing (ST) aims to find the origins of synthetic or manipulated speech. However, ST models must adapt to learn new deepfake attacks while retaining knowledge of the previous ones. A major challenge is catastrophic forgetting, where models lose the ability to recognize previously learned attacks. Some continual learning methods help with deepfake detection, but multi-class tasks such as ST introduce additional challenges as the number of classes grows. To address this, we propose an analytic class incremental learning method called AnaST. When new attacks appear, the feature extractor remains fixed, and the classifier is updated with a closed-form analytical solution in one epoch. This approach ensures data privacy, optimizes memory usage, and is suitable for online training. The experiments carried out in this work show that our method outperforms the baselines.
Recent advances in speech deepfake detection (SDD) have significantly improved artifacts-based detection in spoofed speech. However, most models overlook speech naturalness, a crucial cue for distinguishing bona fide speech from spoofed speech. This study proposes naturalness-aware curriculum learning, a novel training framework that leverages speech naturalness to enhance the robustness and generalization of SDD. This approach measures sample difficulty using both ground-truth labels and mean opinion scores, and adjusts the training schedule to progressively introduce more challenging samples. To further improve generalization, a dynamic temperature scaling method based on speech naturalness is incorporated into the training process. A 23% relative reduction in the EER was achieved in the experiments on the ASVspoof 2021 DF dataset, without modifying the model architecture. Ablation studies confirmed the effectiveness of naturalness-aware training strategies for SDD tasks.
This study explores the potential of using acoustic features of segmental speech sounds to detect deepfake audio. These features are highly interpretable because of their close relationship with human articulatory processes and are expected to be more difficult for deepfake models to replicate. The results demonstrate that certain segmental features commonly used in forensic voice comparison (FVC) are effective in identifying deep-fakes, whereas some global features provide little value. These findings underscore the need to approach audio deepfake detection using methods that are distinct from those employed in traditional FVC, and offer a new perspective on leveraging segmental features for this purpose.
Text-guided diffusion models have significantly advanced image editing, enabling highly realistic and local modifications based on textual prompts. While these developments expand creative possibilities, their malicious use poses substantial challenges for detection of such subtle deepfake edits. To this end, we introduce Explain Edit (X-Edit), a novel method for localizing diffusion-based edits in images. To localize the edits for an image, we invert the image using a pretrained diffusion model, then use these inverted features as input to a segmentation network that explicitly predicts the edited masked regions via channel and spatial attention. Further, we finetune the model using a combined segmentation and relevance loss. The segmentation loss ensures accurate mask prediction by balancing pixel-wise errors and perceptual similarity, while the relevance loss guides the model to focus on low-frequency regions and mitigate high-frequency artifacts, enhancing the localization of subtle edits. To the best of our knowledge, we are the first to address and model the problem of localizing diffusion-based modified regions in images. We additionally contribute a new dataset of paired original and edited images addressing the current lack of resources for this task. Experimental results demonstrate that X-Edit accurately localizes edits in images altered by text-guided diffusion models, outperforming baselines in PSNR and SSIM metrics. This highlights X-Edit's potential as a robust forensic tool for detecting and pinpointing manipulations introduced by advanced image editing techniques.
Advances in AI generative models facilitate super-realistic video synthesis, amplifying misinformation risks via social media and eroding trust in digital content. Several research works have explored new deepfake detection methods on AI-generated images to alleviate these risks. However, with the fast development of video generation models, such as Sora and WanX, there is currently a lack of large-scale, high-quality AI-generated video datasets for forgery detection. In addition, existing detection approaches predominantly treat the task as binary classification, lacking explainability in model decision-making and failing to provide actionable insights or guidance for the public. To address these challenges, we propose \textbf{GenBuster-200K}, a large-scale AI-generated video dataset featuring 200K high-resolution video clips, diverse latest generative techniques, and real-world scenes. We further introduce \textbf{BusterX}, a novel AI-generated video detection and explanation framework leveraging multimodal large language model (MLLM) and reinforcement learning for authenticity determination and explainable rationale. To our knowledge, GenBuster-200K is the {\it \textbf{first}} large-scale, high-quality AI-generated video dataset that incorporates the latest generative techniques for real-world scenarios. BusterX is the {\it \textbf{first}} framework to integrate MLLM with reinforcement learning for explainable AI-generated video detection. Extensive comparisons with state-of-the-art methods and ablation studies validate the effectiveness and generalizability of BusterX. The code, models, and datasets will be released.
With the development of generative artificial intelligence, new forgery methods are rapidly emerging. Social platforms are flooded with vast amounts of unlabeled synthetic data and authentic data, making it increasingly challenging to distinguish real from fake. Due to the lack of labels, existing supervised detection methods struggle to effectively address the detection of unknown deepfake methods. Moreover, in open world scenarios, the amount of unlabeled data greatly exceeds that of labeled data. Therefore, we define a new deepfake detection generalization task which focuses on how to achieve efficient detection of large amounts of unlabeled data based on limited labeled data to simulate a open world scenario. To solve the above mentioned task, we propose a novel Open-World Deepfake Detection Generalization Enhancement Training Strategy (OWG-DS) to improve the generalization ability of existing methods. Our approach aims to transfer deepfake detection knowledge from a small amount of labeled source domain data to large-scale unlabeled target domain data. Specifically, we introduce the Domain Distance Optimization (DDO) module to align different domain features by optimizing both inter-domain and intra-domain distances. Additionally, the Similarity-based Class Boundary Separation (SCBS) module is used to enhance the aggregation of similar samples to ensure clearer class boundaries, while an adversarial training mechanism is adopted to learn the domain-invariant features. Extensive experiments show that the proposed deepfake detection generalization enhancement training strategy excels in cross-method and cross-dataset scenarios, improving the model's generalization.
Advances in computer vision and deep learning have blurred the line between deepfakes and authentic media, undermining multimedia credibility through audio-visual forgery. Current multimodal detection methods remain limited by unbalanced learning between modalities. To tackle this issue, we propose an Audio-Visual Joint Learning Method (MACB-DF) to better mitigate modality conflicts and neglect by leveraging contrastive learning to assist in multi-level and cross-modal fusion, thereby fully balancing and exploiting information from each modality. Additionally, we designed an orthogonalization-multimodal pareto module that preserves unimodal information while addressing gradient conflicts in audio-video encoders caused by differing optimization targets of the loss functions. Extensive experiments and ablation studies conducted on mainstream deepfake datasets demonstrate consistent performance gains of our model across key evaluation metrics, achieving an average accuracy of 95.5% across multiple datasets. Notably, our method exhibits superior cross-dataset generalization capabilities, with absolute improvements of 8.0% and 7.7% in ACC scores over the previous best-performing approach when trained on DFDC and tested on DefakeAVMiT and FakeAVCeleb datasets.
Audio deepfake detection (ADD) has grown increasingly important due to the rise of high-fidelity audio generative models and their potential for misuse. Given that audio large language models (ALLMs) have made significant progress in various audio processing tasks, a heuristic question arises: Can ALLMs be leveraged to solve ADD?. In this paper, we first conduct a comprehensive zero-shot evaluation of ALLMs on ADD, revealing their ineffectiveness in detecting fake audio. To enhance their performance, we propose $\mathcal{A}LLM4ADD$, an ALLM-driven framework for ADD. Specifically, we reformulate ADD task as an audio question answering problem, prompting the model with the question: "Is this audio fake or real?". We then perform supervised fine-tuning to enable the ALLM to assess the authenticity of query audio. Extensive experiments are conducted to demonstrate that our ALLM-based method can achieve superior performance in fake audio detection, particularly in data-scarce scenarios. As a pioneering study, we anticipate that this work will inspire the research community to leverage ALLMs to develop more effective ADD systems.
Deepfake audio detection is challenging for low-resource languages like Bengali due to limited datasets and subtle acoustic features. To address this, we introduce BangalFake, a Bengali Deepfake Audio Dataset with 12,260 real and 13,260 deepfake utterances. Synthetic speech is generated using SOTA Text-to-Speech (TTS) models, ensuring high naturalness and quality. We evaluate the dataset through both qualitative and quantitative analyses. Mean Opinion Score (MOS) from 30 native speakers shows Robust-MOS of 3.40 (naturalness) and 4.01 (intelligibility). t-SNE visualization of MFCCs highlights real vs. fake differentiation challenges. This dataset serves as a crucial resource for advancing deepfake detection in Bengali, addressing the limitations of low-resource language research.
Synthetic media generated by Generative Adversarial Networks (GANs) pose significant challenges in verifying authenticity and tracing dataset origins, raising critical concerns in copyright enforcement, privacy protection, and legal compliance. This paper introduces a novel forensic framework for identifying the training dataset (e.g., CelebA or FFHQ) of GAN-generated images through interpretable feature analysis. By integrating spectral transforms (Fourier/DCT), color distribution metrics, and local feature descriptors (SIFT), our pipeline extracts discriminative statistical signatures embedded in synthetic outputs. Supervised classifiers (Random Forest, SVM, XGBoost) achieve 98-99% accuracy in binary classification (real vs. synthetic) and multi-class dataset attribution across diverse GAN architectures (StyleGAN, AttGAN, GDWCT, StarGAN, and StyleGAN2). Experimental results highlight the dominance of frequency-domain features (DCT/FFT) in capturing dataset-specific artifacts, such as upsampling patterns and spectral irregularities, while color histograms reveal implicit regularization strategies in GAN training. We further examine legal and ethical implications, showing how dataset attribution can address copyright infringement, unauthorized use of personal data, and regulatory compliance under frameworks like GDPR and California's AB 602. Our framework advances accountability and governance in generative modeling, with applications in digital forensics, content moderation, and intellectual property litigation.
We present the first large-scale open-set benchmark for multilingual audio-video deepfake detection. Our dataset comprises over 250 hours of real and fake videos across eight languages, with 60% of data being generated. For each language, the fake videos are generated with seven distinct deepfake generation models, selected based on the quality of the generated content. We organize the training, validation and test splits such that only a subset of the chosen generative models and languages are available during training, thus creating several challenging open-set evaluation setups. We perform experiments with various pre-trained and fine-tuned deepfake detectors proposed in recent literature. Our results show that state-of-the-art detectors are not currently able to maintain their performance levels when tested in our open-set scenarios. We publicly release our data and code at: https://huggingface.co/datasets/unibuc-cs/MAVOS-DD.
The field of Fake Image Detection and Localization (FIDL) is highly fragmented, encompassing four domains: deepfake detection (Deepfake), image manipulation detection and localization (IMDL), artificial intelligence-generated image detection (AIGC), and document image manipulation localization (Doc). Although individual benchmarks exist in some domains, a unified benchmark for all domains in FIDL remains blank. The absence of a unified benchmark results in significant domain silos, where each domain independently constructs its datasets, models, and evaluation protocols without interoperability, preventing cross-domain comparisons and hindering the development of the entire FIDL field. To close the domain silo barrier, we propose ForensicHub, the first unified benchmark & codebase for all-domain fake image detection and localization. Considering drastic variations on dataset, model, and evaluation configurations across all domains, as well as the scarcity of open-sourced baseline models and the lack of individual benchmarks in some domains, ForensicHub: i) proposes a modular and configuration-driven architecture that decomposes forensic pipelines into interchangeable components across datasets, transforms, models, and evaluators, allowing flexible composition across all domains; ii) fully implements 10 baseline models, 6 backbones, 2 new benchmarks for AIGC and Doc, and integrates 2 existing benchmarks of DeepfakeBench and IMDLBenCo through an adapter-based design; iii) conducts indepth analysis based on the ForensicHub, offering 8 key actionable insights into FIDL model architecture, dataset characteristics, and evaluation standards. ForensicHub represents a significant leap forward in breaking the domain silos in the FIDL field and inspiring future breakthroughs.