In this work, we propose a method for estimating depth for an image of a monocular camera in order to avoid a collision for the autonomous flight of a drone. The highest flight speed of a drone is generally approximate 22.2 m/s, and long-distant depth information is crucial for autonomous flights since if the long-distance information is not available, the drone flying at high speeds is prone to collisions. However, long-range, measurable depth cameras are too heavy to be equipped on a drone. This work applies Pix2Pix, which is a kind of Conditional Generative Adversarial Nets (CGAN). Pix2Pix generates depth images from a monocular camera. Additionally, this work applies optical flow to enhance the accuracy of depth estimation. In this work, we propose a highly accurate depth estimation method that effectively embeds an optical flow map into a monocular image. The models are trained with taking advantage of AirSim, which is one of the flight simulators. AirSim can take both monocular and depth images over a hundred meter in the virtual environment, and our model generates a depth image that provides the long-distance information than images captured by a common depth camera. We evaluate accuracy and error of our proposed method using test images in AirSim. In addition, the proposed method is utilized for flight simulation to evaluate the effectiveness to collision avoidance. As a result, our proposed method is higher accuracy and lower error than a state of work. Moreover, our proposed method is lower collision than a state of work.

Existing models achieve high accuracy in small object detection but often overlook the rotational and dense characteristics of targets. To address this, we propose an improved YOLO11-based model specifically designed for detecting dense and rotated small objects in UAV scenarios. To better extract features from rotated targets, we design a C3k2–ARC module that enhances the model’s rotational detection capability. In addition, we introduce the CL-Concat feature fusion module, which combines traditional concatenation with channel and spatial attention, significantly improving the quality of multi-scale feature fusion. The experimental results demonstrate that the proposed method achieves notable performance improvements across multiple public benchmark data sets. Compared with the advanced YOLO11n model, our approach achieves gains of 1.6% on VISDRONE, 1.4% on UAVDT, 0.2% on CARPK, and 0.1% on UAVROD, further validating its effectiveness across diverse UAV detection scenarios.

Recent years, autonomous drones have attracted attention in many fields due to their convenience. Autonomous drones require precise depth information so as to avoid collision to fly fast and both of RGB image and LiDAR point cloud are often employed in applications based on Convolutional Neural Networks (CNNs) to estimate the distance to obstacles. Such applications are implemented onboard embedded systems. In order to precisely estimate the depth, such CNN models are in general so complex to extract many features that the computational complexity increases, requiring long inference time. In order to solve the issue, we employ optical flow to aid in-depth estimation. In addition, we propose a new attention structure that makes maximum use of optical flow without complicating the network. Furthermore, we achieve improved performance without modifying the depth estimator by adding a perceptual discriminator in training. The proposed model is evaluated through accuracy, error, and inference time on the KITTI dataset. In the experiments, we have demonstrated the proposed method achieves better performance by up to 34% accuracy, 55% error reduction and 66% faster inference time on Jetson nano compared to previous methods. The proposed method is also evaluated through a collision avoidance in simulated drone flight and achieves the lowest collision rate of all estimation methods. These experimental results show the potential of proposed method to be used in real-world autonomous drone flight applications.

Self-supervised monocular depth estimation is gaining significant attention because it can learn depth from video without needing expensive ground-truth data. However, many self-supervised models remain too heavy for edge devices, and simply shrinking them tends to degrade accuracy. To address this trade-off, we present MonoLENS, an extension of Lite-Mono. MonoLENS follows a design that reduces computation while preserving geometric fidelity (relative depth relations, boundaries, and planar structures). MonoLENS advances Lite-Mono by suppressing computation on paths with low geometric contribution, focusing compute and attention on layers rich in structural cues, and pruning redundant operations in later stages. Our model incorporates two new modules, the DS-Upsampling Block and the MCACoder, along with a simplified encoder. Specifically, the DS-Upsampling Block uses depthwise separable convolutions throughout the decoder, which greatly lowers floating-point operations (FLOPs). Furthermore, the MCACoder applies Multidimensional Collaborative Attention (MCA) to the output of the second encoder stage, helping to make edge details sharper in high-resolution feature maps. Additionally, we simplified the encoder’s architecture by reducing the number of blocks in its fourth stage from 10 to 4, which resulted in a further reduction of model parameters. When tested on both the KITTI and Cityscapes benchmarks, MonoLENS achieved leading performance. On the KITTI benchmark, MonoLENS reduced the number of model parameters by 42% (1.8M) compared with Lite-Mono, while simultaneously improving the squared relative error by approximately 4.5%.

High-precision 3D object detection is pivotal for autonomous driving. However, voxel-based two-stage detectors still struggle with small and non-rigid objects due to the misalignment between classification confidence and localization accuracy, and the loss of fine-grained spatial context during feature flattening. To address these issues, we propose the Varifocal Fine-grained Refinement Network (VFR-Net). We introduce Varifocal Loss (VFL) to learn IoU-aware scores for prioritizing high-quality proposals, and a Fine-Grained Refinement Attention (FGRA) Module to capture local geometric details via self-attention before flattening. Extensive experiments on the KITTI and ONCE datasets demonstrate that VFR-Net consistently outperforms the Voxel R-CNN baseline, improving the overall mAP by +1.12% on KITTI and +2.63% on ONCE. Specifically, it achieves AP gains of +1.81% and +1.28% for pedestrians and cyclists on KITTI (averaged over Easy/Moderate/Hard), and +6.53% and +1.50% on ONCE (Overall).

In recent years, the demand for both high accuracy and real-time performance in 3D object detection has increased alongside the advancement of autonomous driving technology. While multimodal methods that integrate LiDAR and camera data have demonstrated high accuracy, these methods often have high computational costs and latency. To address these issues, we propose an efficient 3D object detection network that integrates three key components: a DepthWise Lightweight Encoder (DWLE) module for efficient feature extraction, an Efficient LiDAR Image Fusion (ELIF) module that combines channel attention with cross-modal feature interaction, and a Mixture of CNN and Point Transformer (MCPT) module for capturing rich spatial contextual information. Experimental results on the KITTI dataset demonstrate that our proposed method outperforms existing approaches by achieving approximately 0.6% higher 3D mAP, 7.6% faster inference speed, and 17.0% fewer parameters. These results highlight the effectiveness of our approach in balancing accuracy, speed, and model size, making it a promising solution for real-time applications in autonomous driving.