The field of UAV-based 3D perception and localization is moving towards more adaptive and efficient systems. Researchers are exploring biologically inspired approaches, such as active sensing behaviors, to improve odometry accuracy and mapping performance in complex environments. Another trend is the development of spherical robots for mapping applications, which offer unique advantages in hazardous or confined environments. Additionally, there is a growing interest in autonomous corridor-based transport systems for UAVs, which can efficiently navigate and transport payloads in cluttered environments. Innovative solutions are also being proposed for neural 3D object reconstruction using small-scale UAVs, enabling high-fidelity scanning of static objects in constrained environments. Noteworthy papers include: AEOS, which proposes a hybrid architecture combining model predictive control and reinforcement learning for adaptive LiDAR control. Acetrans, which presents a unified perception, planning, and control framework for autonomous UAV suspended transport systems. Neural 3D Object Reconstruction, which introduces a dual-reconstruction pipeline for fully autonomous, high-fidelity 3D scanning of static objects using UAVs weighing under 100 grams. PERAL, which achieves passive LiDAR excitation without dedicated hardware, enriching vertical scan diversity while preserving navigation accuracy.