Human emotion synthesis is a crucial aspect of affective computing. It involves using computational methods to mimic and convey human emotions through various modalities, with the goal of enabling more natural and effective human-computer interactions. Recent advancements in generative models, such as Autoencoders, Generative Adversarial Networks, Diffusion Models, Large Language Models, and Sequence-to-Sequence Models, have significantly contributed to the development of this field. However, there is a notable lack of comprehensive reviews in this field. To address this problem, this paper aims to address this gap by providing a thorough and systematic overview of recent advancements in human emotion synthesis based on generative models. Specifically, this review will first present the review methodology, the emotion models involved, the mathematical principles of generative models, and the datasets used. Then, the review covers the application of different generative models to emotion synthesis based on a variety of modalities, including facial images, speech, and text. It also examines mainstream evaluation metrics. Additionally, the review presents some major findings and suggests future research directions, providing a comprehensive understanding of the role of generative technology in the nuanced domain of emotion synthesis.

The dynamics of a quadrotor exhibit the challenges of underactuation, strong coupling, and external disturbances. In order to handle these difficulties and enable quadrotors to align the six-degree-of-freedom (6-DOF) pose of the tilted narrow gap during the traversing maneuver, this article proposes a strategy based on simultaneous planning and execution (SPAE). In this strategy, the unknown disturbance is considered in both the planning and execution of the trajectory simultaneously. Specifically, the disturbance is estimated online during system execution and then combined with the attitude angle of the narrow gap and maximum thrust to constraint acceleration of trajectory. Given the state constraints and physical limitations in thrust, velocity, and jerk, a minimum time trajectory planning algorithm is developed based on a piecewise cubic function to realize aggressive flight. In this algorithm, the coupled maximum thrust is decoupled into three independent axes. Each axis can then generate time-efficient third-order trajectories that satisfy arbitrary state constraints and dynamic feasibility. The average computing time of the algorithm is less than 150 $\mu$ s on the STM32 embedded platform, which is suitable for quadrotors with limited onboard computing capability. Finally, simulations and outdoor experiments were conducted to verify the effectiveness of the proposed system strategy. The outdoor results demonstrate that the proposed strategy can enable the quadrotor to carry out aggressive flying through a narrow gap in a natural disturbance environment.

Safety is crucial for quadrotor flight, especially when the system suffers one rotor loss and disturbance. To enable quadrotors to survive in these worst cases, this paper proposes a Simultaneous Planning And Execution (SPAE) system strategy. First, the Euler angles under different equilibrium states are solved to ensure the feasibility of the control input when the quadrotors encounters rotor loss and disturbance. Based on the admissible Euler angle solutions, an emergency control strategy is proposed to realize desired angles by recalculating the desired trajectory and the disturbance observer. Meanwhile, this control strategy sacrifices the yaw angle control to handle further underactuation of the system caused by complete rotor failure. In addition to rotor losses, unsafe flight caused by actuator saturation is considered in the SPAE strategy where the planning and execution are simultaneous. As a decision maker, the planning considers the current disturbances and constraints of actuators, then generates a feasible trajectory for controller execution. As an executor of the decision, the controller feedback estimated disturbance information to the trajectory planning, which can help trajectory planning make admissible decisions consistent with the current situation. As a result, the actuator saturation caused by the underestimation of disturbance can be avoided under a feasible regenerated trajectory. Extensive simulations and experiments prove the validity of the proposed SPAE strategy.

This paper investigates the formation control problem of multi-unmanned aerial vehicle (UAV) systems with multi-operating modes. While mode switching enhances the flexibility of multi-UAV systems, it also introduces dynamic model switching behaviors in UAVs. Moreover, obtaining an accurate dynamic model for a multi-UAV system is challenging in practice. In addition, communication link failures and time-varying unknown disturbances are inevitable in multi-UAV systems. Hence, to overcome the adverse effects of the above challenges, a hybrid iterative learning formation control strategy is proposed in this paper. The proposed controller does not rely on precise modeling and exhibits its learning ability by utilizing historical input–output data to update the current control input. Furthermore, two convergence theorems are proven to guarantee the convergence of state, disturbance estimation, and formation tracking errors. Finally, three simulation examples are conducted for a multi-UAV system consisting of four quadrotor UAVs under multi-operating modes, switching topologies, and external disturbances. The results of the simulations show the strategy’s effectiveness and superiority in achieving the desired formation control objectives.

Despite the ubiquity of unmanned aerial vehicles (UAVs) in urban monitoring, nighttime UAV imagery—plagued by high level of noise, uneven light, motion blur, and low dynamic range—poses unique challenges for 3D reconstruction. These degradations cripple traditional Structure from Motion (SfM) frameworks through unreliable feature extraction and inconsistent cross-view spatial association. We propose a 3D reconstruction framework to recover accurate camera poses and point clouds from unordered UAV images captured in nighttime urban scenes. Traditional SfM pipelines typically rely on well-lighted scene imagery, which becomes challenging in low-light conditions due to poor keypoint detection and matching, potentially leading to a breakdown of the entire SfM system. To address these challenges, we introduce a Low-Light Enhancement SfM (LLE-SfM) framework that leverages an enhancement network to suppress noise while restoring scene colors and texture details. Specifically, our framework first decomposes every image of a set of nighttime UAV images into high/low-frequency layers. The High-frequency Stream is employed to remove noise and enhance edge details, while transformer layers are deployed within the Low-frequency Stream to reconstruct scene colors and refine consistency. Due to the lack of low-light 3D reconstruction benchmark datasets for nighttime urban scenes, we introduce the LLUrban3d dataset, including the Rendered subset and Real subset, to provide sufficient ground truth for the training and evaluation of low-light 3D reconstruction pipeline. Experiments on the LLUrban3d dataset demonstrate that our framework significantly improves reconstruction accuracy and completeness over SfM baseline in nighttime urban scenes, with a maximum improvement of 24.17% in F-score.

Existing state estimation methods use a combination of measurement error and fixed gains to update high-order states. However, if these fixed gains are not finetuned, it will inevitably lead to significant phase distortion and violation of calculus relationship among updated states. To handle these issues, this article proposes a high-order state-filtering system that directly uses measurement data, thereby providing precise high-order differential states for potential applications. The estimated highest order derivative is produced by an online planned polynomial that satisfies all state-error constraints. To fulfill calculus relationship among updated states, each low-order state is uniquely determined by the integration of this highest order derivative. To validate the effectiveness of the proposed state-filtering system, it is applied to closed-loop position control of a brushless direct current (BLDC) motor. In this application, its rotor angle is measured by a high-resolution encoder. Compared to a differentiator, a high-gain observer, and a Kalman filter, the proposed method outperforms them by eliminating the need for parameter optimization, being insensitive to the sampling time step, effectively mitigating phase distortion, and strictly satisfying calculus relationship among updated states. Experimental results have verified the superiority and robustness of the proposed method.