Human-Scene Interaction (HSI) seeks to generate realistic human behaviors within complex environments, yet it faces significant challenges in handling long-horizon, high-level tasks and generalizing to unseen scenes. To address these limitations, we introduce FantasyHSI, a novel HSI framework centered on video generation and multi-agent systems that operates without paired data. We model the complex interaction process as a dynamic directed graph, upon which we build a collaborative multi-agent system. This system comprises a scene navigator agent for environmental perception and high-level path planning, and a planning agent that decomposes long-horizon goals into atomic actions. Critically, we introduce a critic agent that establishes a closed-loop feedback mechanism by evaluating the deviation between generated actions and the planned path. This allows for the dynamic correction of trajectory drifts caused by the stochasticity of the generative model, thereby ensuring long-term logical consistency. To enhance the physical realism of the generated motions, we leverage Direct Preference Optimization (DPO) to train the action generator, significantly reducing artifacts such as limb distortion and foot-sliding. Extensive experiments on our custom SceneBench benchmark demonstrate that FantasyHSI significantly outperforms existing methods in terms of generalization, long-horizon task completion, and physical realism.

Recent advances in audio-driven portrait animation have demonstrated impressive capabilities. However, existing methods struggle to align with fine-grained human preferences across multiple dimensions, such as motion naturalness, lip-sync accuracy, and visual quality. This is due to the difficulty of optimizing among competing preference objectives, which often conflict with one another, and the scarcity of large-scale, high-quality datasets with multidimensional preference annotations. To address these, we first introduce Talking-Critic, a multimodal reward model that learns human-aligned reward functions to quantify how well generated videos satisfy multidimensional expectations. Leveraging this model, we curate Talking-NSQ, a large-scale multidimensional human preference dataset containing 410K preference pairs. Finally, we propose Timestep-Layer adaptive multi-expert Preference Optimization (TLPO), a novel framework for aligning diffusion-based portrait animation models with fine-grained, multidimensional preferences. TLPO decouples preferences into specialized expert modules, which are then fused across timesteps and network layers, enabling comprehensive, fine-grained enhancement across all dimensions without mutual interference. Experiments demonstrate that Talking-Critic significantly outperforms existing methods in aligning with human preference ratings. Meanwhile, TLPO achieves substantial improvements over baseline models in lip-sync accuracy, motion naturalness, and visual quality, exhibiting superior performance in both qualitative and quantitative evaluations.

Creating a realistic animatable avatar from a single static portrait remains challenging. Existing approaches often struggle to capture subtle facial expressions, the associated global body movements, and the dynamic background. To address these limitations, we propose a novel framework that leverages a pretrained video diffusion Transformer model to generate high-fidelity, coherent talking portraits with controllable motion dynamics. At the core of our work is a dual-stage audio-visual alignment strategy. In the first stage, we employ a clip-level training scheme to establish coherent global motion by aligning audio-driven dynamics across the entire scene, including the reference portrait, contextual objects, and background. In the second stage, we refine lip movements at the frame level using a lip-tracing mask, ensuring precise synchronization with audio signals. To preserve identity without compromising motion flexibility, we replace the commonly used reference network with a lightweight cross-attention module that effectively maintains facial consistency throughout the video. Furthermore, we integrate a motion intensity modulation module that explicitly controls facial keypoints and body joint trajectories, enabling fine-grained manipulation of portrait movements beyond mere lip motion. Extensive experimental results show that our proposed approach achieves higher quality with better realism, coherence, motion intensity, and identity preservation. Our demo, code, models can be found on this page: https://fantasy-amap.github.io/fantasy-talking/.

In recent years, with the rise of deep learning, it has become a hot research topic to combine time-frequency analysis technology with deep learning to recognize radar signals. For the application of deep learning in radar signal recognition, however, the discovery of adversarial examples poses a tremendous security risk. Based on experiments, it appears that the radar signal recognition model based on the time-frequency image have been shown to be less vulnerable to adversarial attack methods based on time domain. Therefore, we propose a cross-modal attack (CMA). Firstly, we establish a surrogate model architecture locally, including three parts: time-frequency analysis, data quantization, and classifier. Secondly, we train this architecture as a whole and generate adversarial examples utilizing the trained surrogate model architecture parameters and adversarial attack methods. Finally, we carry out the CMA on the radar signal recognition model based on the time-frequency image by adding adversarial perturbations to the original signal. According to experimental results, the CMA can reduce the model recognition accuracy by more than 30%, demonstrating good attack performance, when the perturbation strength is 0.1 and the signal-to-noise ratio is 0 dB.