Marine Video Kit: A New Marine Video Dataset
for Content-based Analysis and Retrieval

Quang-Trung Truong¹          Tuan-Anh Vu¹          Tan-Sang Ha¹          Jakub Lokoč²
         Yue Him Wong Tim³          Ajay Joneja¹          Sai-Kit Yeung¹

¹Hong Kong University of Science and Technology ²FMP, Charles University, Prague, Czech Republic
³Shenzhen University

Several examples of dataset video frames that are extracted with 1fps and their ClipCap descriptions.

Abstract

Effective analysis of unusual domain specific video collections represents an important practical problem, where state-of-the-art general purpose models still face limitations. Hence, it is desirable to design benchmark datasets that challenge novel powerful models for specific domains with additional constraints. It is important to remember that domain specific data may be noisier (e.g., endoscopic or underwater videos) and often require more experienced users for effective search. In this paper, we focus on single-shot videos taken from moving cameras in underwater environments, which constitute a nontrivial challenge for research purposes. The first shard of a new Marine Video Kit dataset is presented to serve for video retrieval and other computer vision challenges. Our dataset is used in a special session during Video Browser Showdown 2023. In addition to basic meta-data statistics, we present several insights based on low-level features as well as semantic annotations of selected keyframes. The analysis also contains experiments showing limitations of respected general purpose models for retrieval. Our dataset and code are publicly available at

Materials

Paper

Dataset

Code

Marine Video Kit dataset - Statistics and Videos

Some marine videos are selected from the dataset.

Total length: 12.38 hours
Number of videos: 1374
Dive sites: 11 countries/ regions

Known-item Search Experiment

        
    @inproceedings{MVK,
    title     = {Marine Video Kit: A New Marine Video Dataset for Content-based Analysis and Retrieval},
    author    = {Truong, Quang-Trung and Vu, Tuan-Anh and Ha, Tan-Sang and Lokoč, Jakub and Tim,
                Yue Him Wong and Joneja, Ajay and Yeung, Sai-Kit},
    publisher = {Springer},
    year      = {2023},
    booktitle = {MultiMedia Modeling - 29th International Conference, {MMM} 2023,
                Bergen, Norway, January 9-12, 2023},
    series    = {Lecture Notes in Computer Science},
}
    }