Human detection in images using deep learning has been a popular research topic in recent years and has achieved remarkable performance. Training a human detection network is useful for first responders to search for trapped victims in debris after a disaster. In this paper, we focus on the detection of such victims using deep learning, and we find that state-of-the-art detection models pre-trained on the well-known COCO dataset fail to detect victims. This is because all the people in the training set are shown in photos of daily life or sports activities, while people in the debris after a disaster usually only have parts of their bodies exposed. In addition, because of the dust, the colors of their clothes or body parts are similar to those of the surrounding debris. Compared with collecting images of common objects, images of disaster victims are extremely difficult to obtain for training. Therefore, we propose a framework to generate harmonious composite images for training. We first paste body parts onto a debris background to generate composite victim images and then use a deep harmonization network to make the composite images look more harmonious. We select YOLOv5l as the most suitable model, and experiments show that using composite images for training improves the AP (average precision) by 19.4% ($15.3\%\to 34.7\%$). Furthermore, using the harmonious images is of great benefit to training a better victim detector, and the AP is further improved by 10.2% ($34.7\%\to 44.9\%$). This research is part of the EU project INGENIOUS. Our composite images and code are publicly available on our website.

Autonomous navigation of drones using computer vision has achieved promising performance. Nano-sized drones based on edge computing platforms are lightweight, flexible, and cheap; thus, they are suitable for exploring narrow spaces. However, due to their extremely limited computing power and storage, vision algorithms designed for high-performance GPU platforms cannot be used for nano-drones. To address this issue, this paper presents a lightweight CNN depth estimation network deployed on nano-drones for obstacle avoidance. Inspired by knowledge distillation (KD), a Channel-Aware Distillation Transformer (CADiT) is proposed to facilitate the small network to learn knowledge from a larger network. The proposed method is validated on the KITTI dataset and tested on a Crazyflie nano-drone with an ultra-low power microprocessor GAP8. This paper also implements a communication pipe so that the collected images can be streamed to a laptop through the on-board Wi-Fi module in real-time, enabling an offline reconstruction of the environment.

Autonomous unmanned aerial vehicle (UAV) technology is a promising technology for minimizing human involvement in dangerous activities like urban search and rescue missions (USAR), both in indoor and outdoor. Automated navigation from outdoor to indoor environments is not trivial, as it encompasses the ability of a UAV to automatically map and locate the openings in a damaged building. This study focuses on developing a deep learning model for the detection of damaged building openings in real time. A novel damaged building-opening dataset containing images and mask annotations, as well as a comparison between single and multi-task learning-based detectors are given. The deep learning-based detector used in this study is based on YOLOv5. First, this study compared the different versions of YOLOv5 (i.e., small, medium, and large) capacity to perform damaged building-opening detections. Second, a multitask learning YOLOv5 was trained on the same dataset and compared with the single-task detector. The multitask learning (MTL) was developed based on the YOLOv5 object detection architecture, adding a segmentation branch jointly with the detection head. This study found that the MTL-based YOLOv5 can improve detection performance by combining detection and segmentation losses. The YOLOv5s-MTL trained on the damaged building-opening dataset obtained 0.648 mAP, an increase of 0.167 from the single-task-based network, while its inference speed was 73 frames per second on the tested platform.

Self-supervised monocular depth estimation that does not require ground truth for training has attracted attention in recent years. It is of high interest to design lightweight but effective models so that they can be deployed on edge devices. Many existing architectures benefit from using heavier backbones at the expense of model sizes. This paper achieves comparable results with a lightweight architecture. Specifically, the efficient combination of CNNs and Transformers is investigated, and a hybrid architecture called Lite-Mono is presented. A Consecutive Dilated Convolutions (CDC) module and a Local-Global Features Interaction (LGFI) module are proposed. The former is used to extract rich multi-scale local features, and the latter takes advantage of the self-attention mechanism to encode long-range global information into the features. Experiments demonstrate that Lite-Mono outperforms Monodepth2 by a large margin in accuracy, with about 80% fewer trainable parameters.