Summary of Vessel Re-identification and Activity Detection in Thermal Domain For Maritime Surveillance, by Yasod Ginige et al.
Vessel Re-identification and Activity Detection in Thermal Domain for Maritime Surveillance
by Yasod Ginige, Ransika Gunasekara, Darsha Hewavitharana, Manjula Ariyarathne, Ranga Rodrigo, Peshala Jayasekara
First submitted to arxiv on: 12 Jun 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary A novel thermal, vision-based approach for maritime surveillance is introduced, addressing challenges like vessel re-identification at night and detecting suspicious activities. The method proposes a viewpoint-independent algorithm for vessel re-identification using separate side-spaces and shape information. Techniques are developed to adapt tracking and activity detection algorithms for the thermal domain, trained on a newly created dataset. This dataset becomes the first publicly available benchmark for thermal maritime surveillance. The system achieves 81.8% Top1 score in vessel re-identification and 72.4% frame mAP score in identifying suspicious activities, setting new benchmarks for each task in the thermal domain. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Maritime surveillance is important to prevent illegal activities like drug smuggling and human trafficking. At night, it’s hard to see vessels and detect suspicious activity because of visibility issues. This paper presents a new way to do maritime surveillance using heat sensors and cameras. It helps track objects, identify ships again, and find suspicious behavior. The method uses shape information from the sides of ships to re-identify them, even when they’re not brightly lit. The system is trained on a special dataset created just for this problem. It can accurately identify vessels (81.8%) and detect suspicious activities (72.4%). This is an important step forward in maritime surveillance. |
Keywords
» Artificial intelligence » Tracking
