KOLOMVERSE: Korea Open Large-Scale Image Dataset for Object Detection in the Maritime Universe
Abhilasha Nanda, Sung Won Cho, Hyeopwoo Lee, Jin Hyoung Park
Abstract
Object detection in the maritime domain is crucial for ensuring the safety and navigation of ships. However, there is still a lack of publicly available large-scale datasets in this domain. To address this challenge, we introduce KOLOMVERSE, an open large-scale image dataset for object detection in the maritime domain. We gathered 5,845 hours of video data captured from 21 territorial waters of South Korea. Through rigorous data quality assessment, we manually gathered around 186,419 4K resolution images from the video data. The KOLOMVERSE includes five classes (ship, buoy, fishnet buoy, lighthouse and wind farm) for maritime object detection. The dataset comprises images with dimensions of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$3840 \times 2160$ </tex-math></inline-formula> pixels and, to the best of our knowledge, it is by far the largest publicly available dataset for object detection in the maritime domain. We conducted object detection experiments and evaluated our dataset using several state-of-the-art pre-trained architectures to demonstrate its effectiveness and usefulness. The dataset is available at: <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/MaritimeDataset/KOLOMVERSE</uri>.