A Dataset and Deep Learning Benchmark for Event Recognition in Aerial Videos


Along with the now widespread availability of unmanned aerial vehicles (UAVs), large volumes of aerial videos have been produced. It is unrealistic for humans to screen such big data and understand their contents. Hence methodological research on the automatic understanding of UAV videos is of paramount importance. In this work, we introduce a novel task of event recognition in unconstrained aerial videos in the remote sensing community and present a large-scale, human-annotated dataset, named ERA (Event Recognition in Aerial videos), consisting of 2,864 videos each with a label from 25 different classes corresponding to an event unfolding 5 seconds. The ERA dataset is designed to have a significant intra-class variation and inter-class similarity and captures dynamic events in different environments and at different scales. Moreover, in order to provide a benchmark for this task, we extensively evaluate existing deep networks. We expect that the ERA dataset will facilitate further progress in automatic aerial video comprehension.


The goal of this work is to collect a large, diverse dataset for training models for event understanding in UAV videos. As we gather aerial videos from Youtube, the largest video sharing platform in the world, we are capable of including a large breadth of diversity that would be more challenging than making use of self-collected data. In total, we have gathered and annotated 2,864 videos for 25 classes. Each video sequence is at 24 fps (frames per second), in 5 seconds, and with a spatial size of 640×640 pixels.



1. Dataset

Download Link

2. Trained Models

Single-frame Classification Models

Model Weights Download Model Weights Download
VGG-16 GoogleDrive BaiduPan MobileNet GoogleDrive BaiduPan
VGG-19 GoogleDrive BaiduPan DenseNet-121 GoogleDrive BaiduPan
Inception-v3 GoogleDrive BaiduPan DenseNet-169 GoogleDrive BaiduPan
ResNet-50 GoogleDrive BaiduPan DenseNet-201 GoogleDrive BaiduPan
ResNet-101 GoogleDrive BaiduPan NASNet-L GoogleDrive BaiduPan
ResNet-152 GoogleDrive BaiduPan

Video Classification Models

Model Weights Download Model Weights Download
C3D-Sport1M GoogleDrive BaiduPan I3D-Kinetics GoogleDrive BaiduPan
C3D-UCF101 GoogleDrive BaiduPan I3D-Kinetics+ImageNet GoogleDrive BaiduPan
P3D-Kinetics GoogleDrive BaiduPan TRN-Something-Something-V2 GoogleDrive BaiduPan
P3D-Kinetics-600 GoogleDrive BaiduPan TRN-Moments-In-Time GoogleDrive BaiduPan

Note: Hovering your mouse over "BaiduPan" to see the extraction code (提取码)

3. Citation

    If you use this dataset for your work, please use the following citation:

        title = {{ERA: A dataset and deep learning benchmark for event recognition in aerial videos}},
        author = {Mou, L. and Hua, Y. and Jin, P. and Zhu, X. X.},
        journal = {IEEE Geoscience and Remote Sensing Magazine},
        year = {in press}