Despite the recent progress of audio-driven video generation, existing methods mostly focus on driving facial movements, leading to non-coherent head and body dynamics. Moving forward, it is desirable yet challenging to generate holistic human videos with both accurate lip-sync and delicate co-speech gestures w.r.t. given audio. In this work, we propose AudCast, a generalized audio-driven human video generation framework adopting a cascade Diffusion-Transformers (DiTs) paradigm, which synthesizes holistic human videos based on a reference image and a given audio. 1) Firstly, an audio-conditioned Holistic Human DiT architecture is proposed to directly drive the movements of any human body with vivid gesture dynamics. 2) Then to enhance hand and face details that are well-knownly difficult to handle, a Regional Refinement DiT leverages regional 3D fitting as the bridge to reform the signals, producing the final results. Extensive experiments demonstrate that our framework generates high-fidelity audio-driven holistic human videos with temporal coherence and fine facial and hand details. Resources can be found at https://guanjz20.github.io/projects/AudCast.

The non-consensual exploitation of facial manipulation has emerged as a pressing societal concern. In tandem with the identification of such fake content, recent research endeavors have advocated countering manipulation techniques through proactive interventions, specifically the incorporation of adversarial noise to impede the manipulation in advance. Nevertheless, with insufficient consideration of robustness, we show that current methods falter in providing protection after simple perturbations, e.g., blur. In addition, traditional optimization-based methods face limitations in scalability as they struggle to accommodate the substantial expansion of data volume, a consequence of the time-intensive iterative pipeline. To solve these challenges, we propose a learning-based model, Adversarial Robust Safeguard (ARS), to generate desirable protection noise in a single forward process, concurrently exhibiting a heightened resistance against prevalent perturbations. Specifically, our method involves a two-way protection design, characterized by a basic protection component responsible for generating efficacious noise features, coupled with robust protection for further enhancement. In robust protection, we first fuse image features with spatially duplicated noise embedding, thereby accounting for inherent information redundancy. Subsequently, a combination comprising a differentiable perturbation module and an adversarial network is devised to simulate potential information degradation during the training process. To evaluate it, we conduct experiments on four manipulation methods and compare recent works comprehensively. The results of our method exhibit good visual effects with pronounced robustness against varied perturbations at different levels.

Although significant progress has been made in human video generation, most previous studies focus on either human facial animation or full-body animation, which cannot be directly applied to produce realistic conversational human videos with frequent hand gestures and various facial movements simultaneously. To address these limitations, we propose a 2D human video generation framework, named ShowMaker, capable of generating high-fidelity half-body conversational videos via fine-grained diffusion modeling. We leverage dual-stream diffusion models as the backbone of our framework and carefully design two novel components for crucial local regions (i.e., hands and face) that can be easily integrated into our backbone. Specifically, to handle the challenging hand generation caused by sparse motion guidance, we propose a novel Key Point-based Fine-grained Hand Modeling module by amplifying positional information from raw hand key points and constructing a corresponding key point-based codebook. Moreover, to restore richer facial details in generated results, we introduce a Face Recapture module, which extracts facial texture features and global identity features from the aligned human face and integrates them into the diffusion process for face enhancement. Extensive quantitative and qualitative experiments demonstrate the superior visual quality and temporal consistency of our method.

Despite recent advances in syncing lip movements with any audio waves, current methods still struggle to balance generation quality and the model's generalization ability. Previous studies either require long-term data for training or produce a similar movement pattern on all subjects with low quality. In this paper, we propose StyleSync, an effective framework that enables high-fidelity lip synchronization. We identify that a style-based generator would sufficiently enable such a charming property on both one-shot and few-shot scenarios. Specifically, we design a mask-guided spatial information encoding module that preserves the details of the given face. The mouth shapes are accurately modified by audio through modulated convolutions. Moreover, our design also enables personalized lip-sync by introducing style space and generator refinement on only limited frames. Thus the identity and talking style of a target person could be accurately preserved. Extensive experiments demonstrate the effectiveness of our method in producing high-fidelity results on a variety of scenes. Resources can be found at https:/hangz-nju-cuhk.github.io/projects/StyleSync.

Abuses of forgery techniques have created a considerable problem of misinformation on social media. Although scholars devote many efforts to face forgery detection (a.k.a DeepFake detection) and achieve some results, two issues still hinder the practical application. 1) Most detectors do not generalize well to unseen datasets. 2) In a supervised manner, most previous works require a considerable amount of manually labeled data. To address these problems, we propose a simple contrastive pertaining framework for DeepFake detection (DFCP), which works in a finetuning-after-pretraining manner, and requires only a few labels (5%). Specifically, we design a two-stream framework to simultaneously learn high-frequency texture features and high-level semantics information during pretraining. In addition, a video-based frame sampling strategy is proposed to mitigate potential noise data in the instance-discriminative contrastive learning to achieve better performance. Experimental results on several downstream datasets show the state-of-the-art performance of the proposed DFCP, which works at frame-level (w/o temporal reasoning) with high efficiency but outperforms video-level methods.

The emergence of photo-realistic deepfakes on a large scale has become a significant societal concern, which has garnered considerable attention from the research community. Several recent studies have identified the critical issue of “temporal inconsistency” resulting from the frame reassembling process of deepfake generation techniques. However, due to the lack of task-specific design, the spatio-temporal modeling of current methods remains insufficient in three critical aspects: 1) inapparent temporal changes are prone to be undermined compared to abundant spatial cues; 2) minor inconsistent regions are often concealed by motions with greater amplitude during downsampling; 3) capturing both transient inconsistencies and persistent motions simultaneously remains a significant challenge. In this paper, we propose a novel Dual-Modality Co-Learning framework tailored for these characteristics, which achieves more effectual deepfake detection with complementary information from RGB and optical flow modalities. In particular, we designed a Multi-Scale Motion Regularization module to encourage the network to equally prioritize both the significant spatial cues and the subtle temporal facial motion cues. Additionally, we developed a Multi-Span Cross-Attention module to effectively integrate the information from both RGB and optical flow modalities and improve the detection accuracy with multi-span predictions. Extensive experiments validate the effectiveness our ideas and demonstrate the superior performance of our approach.

Recent advances in face forgery techniques produce nearly visually untraceable deepfake videos, which could be leveraged with malicious intentions. As a result, researchers have been devoted to deepfake detection. Previous studies have identified the importance of local low-level cues and temporal information in pursuit to generalize well across deepfake methods, however, they still suffer from robustness problem against post-processings. In this work, we propose the Local- & Temporal-aware Transformer-based Deepfake Detection (LTTD) framework, which adopts a local-to-global learning protocol with a particular focus on the valuable temporal information within local sequences. Specifically, we propose a Local Sequence Transformer (LST), which models the temporal consistency on sequences of restricted spatial regions, where low-level information is hierarchically enhanced with shallow layers of learned 3D filters. Based on the local temporal embeddings, we then achieve the final classification in a global contrastive way. Extensive experiments on popular datasets validate that our approach effectively spots local forgery cues and achieves state-of-the-art performance.