The Regularizing Power of Language-Training Deepfake Detectors

Recently, thanks to the advent of Multimodal-LLMs, deepfake detectors are striving not only to be generalizable but also interpretable. We propose that these two challenges can effectively be tackled jointly, since describable artifacts typically generalize better, opening the possibility to use language as a regularization mechanism. Since deepfake detection generally suffers from overfitting to low-level domain-specific artifacts, our intuition is that an LLM that has been pretrained on language would prefer high-level artifacts that can be described better. This way, we can use high-level features where possible, while training the model to use low-level features where necessary. We utilize a dual-encoder architecture, pairing a frozen specialist detector with a LoRA-tuned MLLM encoder, and a two-stage training curriculum: first, a binary alignment phase demonstrates that the intrinsic capability of MLLMs can effectively combine features to mitigate overfitting to dataset-specific artifacts. To further bolster generalization and achieve interpretability, we employ a reinforcement learning stage that encourages the model to generate descriptive reasoning before classifying, using only binary labels. By rewarding this "explain-then-classify" behavior, we explicitly incentivize the model to prioritize high-level, robust features. Crucially, this process yields both interpretable descriptions and a further boost in cross-dataset performance, even when reasoning chains are omitted at inference. Extensive experiments on benchmark datasets validate our approach, outperforming state-of-the-art methods by a large margin.

Robust Deepfake Detection, NTIRE 2026 Challenge: Report

Robustness is a long-overlooked problem in deepfake detection. However, detection performance is nearly worthless in the real world if it suffers under exposure to even slight image degradation. In addition to weaker degradations that can accidentally occur in the image processing pipeline, there is another risk of malicious deepfakes that specifically introduce degradations, purposefully exploiting the detector's weaknesses in that regard. Here, we present an overview of the NTIRE 2026 Robust Deepfake Detection Challenge, which specifically addresses that problem. Participants were tasked with building a detector that would later be tested on an unknown test-set, which included both common and uncommon degradations of various strengths. With a total number of 337 participants and 57 submissions to the final leaderboard, the first edition of the challenge was well received. To ensure the reliability of the results, participants were given only 24h to complete the test run with no labels provided, limiting the possibility of training on the test data. Furthermore, the top solutions were scored on a private test-set to detect any such overfitting. This report presents the competition setting, dataset preparation, as well as details and performance of methods. Top methods rely on large foundation models, ensembles, and degradation training to combine generality and robustness.

Practical Manipulation Model for Robust Deepfake Detection

Modern deepfake detection models have achieved strong performance even on the challenging cross-dataset task. However, detection performance under non-ideal conditions remains very unstable, limiting success on some benchmark datasets and making it easy to circumvent detection. Inspired by the move to a more real-world degradation model in the area of image super-resolution, we have developed a Practical Manipulation Model (PMM) that covers a larger set of possible forgeries. We extend the space of pseudo-fakes by using Poisson blending, more diverse masks, generator artifacts, and distractors. Additionally, we improve the detectors' generality and robustness by adding strong degradations to the training images. We demonstrate that these changes not only significantly enhance the model's robustness to common image degradations but also improve performance on standard benchmark datasets. Specifically, we show clear increases of $3.51\%$ and $6.21\%$ AUC on the DFDC and DFDCP datasets, respectively, over the s-o-t-a LAA backbone. Furthermore, we highlight the lack of robustness in previous detectors and our improvements in this regard. Code can be found at https://github.com/BenediktHopf/PMM