With the rapid advancements in diffusion models and 3D generation techniques, dynamic 3D content generation has become a crucial research area. However, achieving high-fidelity 4D (dynamic 3D) generation with strong spatial-temporal consistency remains a challenging task. Inspired by recent findings that pretrained diffusion features capture rich correspondences, we propose FB-4D, a novel 4D generation framework that integrates a Feature Bank mechanism to enhance both spatial and temporal consistency in generated frames. In FB-4D, we store features extracted from previous frames and fuse them into the process of generating subsequent frames, ensuring consistent characteristics across both time and multiple views. To ensure a compact representation, the Feature Bank is updated by a proposed dynamic merging mechanism. Leveraging this Feature Bank, we demonstrate for the first time that generating additional reference sequences through multiple autoregressive iterations can continuously improve generation performance. Experimental results show that FB-4D significantly outperforms existing methods in terms of rendering quality, spatial-temporal consistency, and robustness. It surpasses all multi-view generation tuning-free approaches by a large margin and achieves performance on par with training-based methods. Our code and data will be publicly available to support future research.

In this paper, we focus on the task of conditional image generation, where an image is synthesized according to user instructions. The critical challenge underpinning this task is ensuring both the fidelity of the generated images and their semantic alignment with the provided conditions. To tackle this issue, previous studies have employed supervised perceptual losses derived from pre-trained models, i.e., reward models, to enforce alignment between the condition and the generated result. However, we observe one inherent shortcoming: considering the diversity of synthesized images, the reward model usually provides inaccurate feedback when encountering newly generated data, which can undermine the training process. To address this limitation, we propose an uncertainty-aware reward modeling, called Ctrl-U, including uncertainty estimation and uncertainty-aware regularization, designed to reduce the adverse effects of imprecise feedback from the reward model. Given the inherent cognitive uncertainty within reward models, even images generated under identical conditions often result in a relatively large discrepancy in reward loss. Inspired by the observation, we explicitly leverage such prediction variance as an uncertainty indicator. Based on the uncertainty estimation, we regularize the model training by adaptively rectifying the reward. In particular, rewards with lower uncertainty receive higher loss weights, while those with higher uncertainty are given reduced weights to allow for larger variability. The proposed uncertainty regularization facilitates reward fine-tuning through consistency construction. Extensive experiments validate the effectiveness of our methodology in improving the controllability and generation quality, as well as its scalability across diverse conditional scenarios, including segmentation mask, edge, and depth conditions.

Video identity customization seeks to synthesize realistic, temporally coherent videos of a specific subject, given a single reference image and a text prompt. This task presents two core challenges: (1) maintaining identity consistency while aligning with the described appearance and actions, and (2) generating natural, fluid motion without unrealistic stiffness. To address these challenges, we introduce Proteus-ID, a novel diffusion-based framework for identity-consistent and motion-coherent video customization. First, we propose a Multimodal Identity Fusion (MIF) module that unifies visual and textual cues into a joint identity representation using a Q-Former, providing coherent guidance to the diffusion model and eliminating modality imbalance. Second, we present a Time-Aware Identity Injection (TAII) mechanism that dynamically modulates identity conditioning across denoising steps, improving fine-detail reconstruction. Third, we propose Adaptive Motion Learning (AML), a motion-aware optimization strategy that reweights training loss based on optical-flow-derived motion heatmaps, enhancing motion realism without requiring additional inputs. To support this task, we construct Proteus-Bench, a high-quality dataset comprising 200K curated clips for training and 150 individuals from diverse professions and ethnicities for evaluation. Extensive experiments demonstrate that Proteus-ID outperforms prior methods in identity preservation, text alignment, and motion quality, establishing a new benchmark for video identity customization.

Visual anagrams are images that change appearance upon transformation, like flipping or rotation. With the advent of diffusion models, generating such optical illusions can be achieved by averaging noise across multiple views during the reverse denoising process. However, we observe two critical failure modes in this approach: (i) concept segregation, where concepts in different views are independently generated, which can not be considered a true anagram, and (ii) concept domination, where certain concepts overpower others. In this work, we cast the visual anagram generation problem in a multi-task learning setting, where different viewpoint prompts are analogous to different tasks, and derive denoising trajectories that align well across tasks simultaneously. At the core of our designed framework are two newly introduced techniques, where (i) an anti-segregation optimization strategy that promotes overlap in cross-attention maps between different concepts, and (ii) a noise vector balancing method that adaptively adjusts the influence of different tasks. Additionally, we observe that directly averaging noise predictions yields suboptimal performance because statistical properties may not be preserved, prompting us to derive a noise variance rectification method. Extensive qualitative and quantitative experiments demonstrate our method's superior ability to generate visual anagrams spanning diverse concepts.

In the Internet of Vehicles (IoV), task offloading is crucial for enhancing computational efficiency and managing resource constraints. However, the presence of information asymmetry and the inherent selfish behavior of vehicles pose significant challenges to task offloading in IoV. Thus, to improve the reliability and efficiency of offloading in scenarios with information asymmetry, an innovative two-stage offloading strategy that employs reinforcement learning to optimize task allocation and scheduling is introduced in this paper. To be specific, the first stage in the proposed offloading strategy involves an initial task allocation strategy based on task attributes, the state of the intermediate buffer queue, and the maximum task reception capacity of Roadside Units (RSUs). The second stage explores partial offloading strategies, investigating the impact of intermediate incentives on offloading decisions. By focusing on maximizing social utility, the destinations, offloading rates, and advantageous portions of intermediate incentives for partial offloading tasks are determined in this paper as well. Moreover, the Proximal Policy Optimization (PPO) and Soft Actor-Critic (SAC) algorithms are used to solve the optimization problem efficiently. Finally, simulation results demonstrate that the proposed innovative strategy significantly outperforms traditional approaches in terms of task completion reliability and efficiency, providing a robust solution for IoV scenarios with information asymmetry.

Recent studies leverage time-aware shaper (TAS) and frame replication and elimination for reliability (FRER) techniques to achieve deterministic latency and high reliability for time-triggered (TT) flows. FRER requires transmitting TT flows on$k$disjoint paths to tolerate transient and permanent failures. However, directly allocating timeslot resources for all$k$TT flows will dramatically increase the computational complexity of gate control lists (GCLs), seriously impair scheduling capabilities, and result in a wastage of bandwidth. In this paper, we propose an ultra-reliable time-aware shaper (uTAS). uTAS only allocates timeslots for one TT flow to ensure deterministic transmission. The$k - 1$replica TT (RT) flows are delivered using a best-effort strategy. On this basis, we propose an adaptive window scheduling (AWS) algorithm based on network calculus, which aims to guarantee that RT flows reach their destinations within the deadline. Evaluation results show that uTAS can meet the flow reliability and deadline requirements. Compared to directly combining TAS and FRER (DCTF), uTAS reduces the total number of GCLs by approximately 72.9%.

With the fast development of the Internet of Vehicles (IoV), the determination of task offloading destinations has been extensively studied. However, the lack of global information makes the existing research unable to apply in the scenario of information asymmetry. In this paper, we first propose a novel intermediary vehicle-assisted task offloading (IVATO) framework under information asymmetry. Then, we develop an intermediary vehicle election mechanism comprehensively considering the trust value and information mastery degree. In this mechanism, we develop a trust value evaluation method based on resilience. In addition, we introduce the indicator of information mastery degree to measure how much information the vehicle has. Finally, simulation results manifest that the proposed trust evaluation method is more reasonable in the long term compared to existing works. Moreover, the performance of the intermediary vehicles can be stably maintained between 0.85 and 1.