This paper proposes an adaptive stabilization control mechanism by using deep reinforcement learning (DRL) for hovering drones that have to execute a surveillance task for a long time. For long-endurance flights, we design and implement a buoyancy-aided autonomous aerial vehicle (AAV) that can use buoyancy lift to decrease the weight and increase the battery capacity so that the flight time can be significantly extended. However, the balloons of the buoyancy-aided AAV can cause “an inverted pendulum effect” and an instability issue on the drone attitude because the increased surface is easily affected by the gusty wind. We propose a buoyancy-aided adaptive stabilization control (BAASC) method with the DRL to stabilize the attitude and extend the flight time of the quadrotor-based buoyancy-aided AAV. This proposed model can immediately control the speeds of all rotors to balance the attitude based on the current state of the drone. Therefore, the degree of swing can be stabilized, and the inverted pendulum effect can be eliminated. The experimental results reveal that the designed buoyancy-aided AAV with the proposed BAASC scheme can effectively stabilize the attitude to extend the flight time by 112.8% compared with a nonbuoyancy-aided AAV under a gusty wind disturbance.

Expanded complementary writing reflecting on the multi-output PAR project, including insights assembled and gleaned for a presentation for Storytellers and Machines (2024 conference at Manchester Metropolitan University) and the overall, 3-stage process of BOBBY....'s development (2019, 2021 and 2024).

A human action recognition system based on image depth is proposed in this paper. Depth information features are not easily disturbed by noise; and due to this characteristic, the system can quickly extract foreground targets. Moreover, the target data, namely, depth and two-dimensional (2D) data, are projected to three orthogonal planes. In this manner, the action featured in the depth motion along the optical axis can clearly describe the trajectory. Based on the change of motion energy and the angle variations of motion orientations, the temporal segmentation (TS) method automatically segments the complex action into several simple movements. Three-dimensional (3D) data is further applied to acquire the three-viewpoint (3V) motion history trajectory, whereby a target’s motion is described through the motion history images (MHIs) from the 3Vs. The weightings corresponding to the gradients of the MHIs are included for determining the viewpoint that bests describe the target’s motion. In terms of feature extraction, the application of multi-resolution motion history histograms can effectively reduce the computational load and achieve a high recognition rate. Experimental results demonstrate that the proposed method can effectively solve the self-occlusion problem.

In this paper, we present a simple but fast codebook generation algorithm, called PREGLA (pattern reduction enhanced GLA). The proposed algorithm is fundamentally different from the previous approaches in that the latter focuses on reducing the size of the codebook whereas the former focuses on using pattern reduction to reduce the computation time. The proposed algorithm is motivated by the observation that input vectors that are ldquostaticrdquo during the training process can be considered as part of the final solutions and thus can be compressed and removed to eliminate the redundant computations at the later iterations of the training process. To evaluate the performance of the proposed algorithm, we compare the proposed algorithm with ldquoefficientrdquo VQ algorithms published recently such as nearest partition set search, fast vector quantization algorithm, and standard GLA. Our experimental results indicate that the proposed algorithm can reduce the computation time from 27.86% up to about 80.45% compared to that of standard GLA and other fast GLA-based algorithms alone.