Anomaly detection is an area of video analysis and plays an increasing role in ensuring safety, preventing risks, and guaranteeing quick response in intelligent surveillance systems. It has become a popular research topic and has piqued the interest of researchers in different communities, such as computer vision, machine learning, remote sensing, and data mining, in recent years. This promotes novel mobile systems where drones are equipped with cameras to help people find better and more efficient solutions to automatically detect anomalies (e.g., car accidents, traffic congestion, street fighting) in traffic surveillance videos. However, anomaly detection methods are still rarely studied and developed in the remote sensing community due to anomalous events rarely occurring in real life, along with the high similarities between the objects of interest with small sizes, multi-scale objects, complex backgrounds of great variations, and high overlap between objects. Therefore, in order to fully exploit the spatio-temporal information for anomaly detection in traffic surveillance circumstances, we propose a future frame prediction network based on transformer architectures to detect abnormal events from drone videography in an unsupervised way. Our model treats consecutive video frames from an input clip and feeds features to a transformer encoder to capture spatial and temporal representations from the sequence. Then, it leverages a decoder to predict the next frame. Furthermore, an event with high reconstruction error is identified as an anomaly in the test phase. Thoroughly empirical studies demonstrate that our method achieves superior performance on the UIT-ADrone dataset and largely outperforms the state-of-the-art anomaly methods on the Drone-Anomaly dataset in aerial surveillance. The source code is available online at https://github.com/Tungufm/ASTT.

Anomaly detection plays an increasingly important role in video surveillance and is one of the issues that has attracted various communities, such as computer vision, machine learning, and data mining in recent years. Moreover, drones equipped with cameras have quickly been deployed to a wide range of applications, starting from border security applications to street monitoring systems. However, there is a notable lack of adequate drone-based datasets available to detect unusual events in the urban traffic environment, especially in roundabouts, due to the density of interaction between road users and vehicles. To promote the development of anomalous event detection with drones in the complex traffic environment, we construct a novel large-scale drone dataset to detect anomalies involving realistic roundabouts in Vietnam, covering a large variety of anomalous events. Traffic at a total of three different roundabouts in Hochiminh city was recorded from a camera-equipped drone. The resulting dataset contains 51 videos with total data traffic of nearly 6.5 hours, captured across 206K frames with ten abnormal event types. Based on this dataset, we comprehensively evaluate the current state-of-the-art algorithms and what anomaly detection can do in drone-based video surveillance. This study presents a detailed description of the proposed UIT-ADrone dataset, along with information regarding data distribution, protocols for evaluation, baseline experimental results on our dataset and other benchmark datasets, discussions, and paves the way for future work. The dataset is available online for non-commercial research at https://uit-together.github.io/datasets/UIT-ADrone/ .

Anomaly analysis is an important component of any surveillance system. In recent years, it has drawn the attention of the computer vision and machine learning communities. In this article, our overarching goal is thus to provide a coherent and systematic review of state-of-the-art techniques and a comprehensive review of the research works in anomaly analysis. We will provide a broad vision of computational models, datasets, metrics, extensive experiments, and what anomaly analysis can do in images and videos. Intensively covering nearly 200 publications, we review (i) anomaly related surveys, (ii) taxonomy for anomaly problems, (iii) the computational models, (iv) the benchmark datasets for studying abnormalities in images and videos, and (v) the performance of state-of-the-art methods in this research problem. In addition, we provide insightful discussions and pave the way for future work.

Unmanned aircraft systems or drones enable us to record or capture many scenes from the bird’s-eye view and they have been fast deployed to a wide range of practical domains, i.e., agriculture, aerial photography, fast delivery and surveillance. Object detection task is one of the core steps in understanding videos collected from the drones. However, this task is very challenging due to the unconstrained viewpoints and low resolution of captured videos. While deep-learning modern object detectors have recently achieved great success in general benchmarks, i.e., PASCAL-VOC and MS-COCO, the robustness of these detectors on aerial images captured by drones is not well studied. In this paper, we present an evaluation of state-of-the-art deep-learning detectors including Faster R-CNN (Faster Regional CNN), RFCN (Region-based Fully Convolutional Networks), SNIPER (Scale Normalization for Image Pyramids with Efficient Resampling), Single-Shot Detector (SSD), YOLO (You Only Look Once), RetinaNet, and CenterNet for the object detection in videos captured by drones. We conduct experiments on VisDrone2019 dataset which contains 96 videos with 39,988 annotated frames and provide insights into efficient object detectors for aerial images.

The application of chest X-ray imaging for early disease screening is attracting interest from the computer vision and deep learning community. To date, various deep learning models have been applied in X-ray image analysis. However, models perform inconsistently depending on the dataset. In this paper, we consider each individual model as a medical doctor. We then propose a doctor consultation-inspired method that fuses multiple models. In particular, we consider both early and late fusion mechanisms for consultation. The early fusion mechanism combines the deep learned features from multiple models, whereas the late fusion method combines the confidence scores of all individual models. Experiments on two X-ray imaging datasets demonstrate the superiority of the proposed method relative to baseline. The experimental results also show that early consultation consistently outperforms the late consultation mechanism in both benchmark datasets. In particular, the early doctor consultation-inspired model outperforms all individual models by a large margin, i.e., 3.03 and 1.86 in terms of accuracy in the UIT COVID-19 and chest X-ray datasets, respectively.