-
[pdf]
[bibtex]@InProceedings{Kiefer_2024_WACV, author = {Kiefer, Benjamin and \v{Z}ust, Lojze and Kristan, Matej and Per\v{s}, Janez and Ter\v{s}ek, Matija and Wiliem, Arnold and Messmer, Martin and Yang, Cheng-Yen and Huang, Hsiang-Wei and Jiang, Zhongyu and Kuo, Heng-Cheng and Mei, Jie and Hwang, Jenq-Neng and Stadler, Daniel and Sommer, Lars and Huang, Kaer and Zheng, Aiguo and Chong, Weitu and Lertniphonphan, Kanokphan and Xie, Jun and Chen, Feng and Li, Jian and Wang, Zhepeng and Zedda, Luca and Loddo, Andrea and Di Ruberto, Cecilia and Vu, Tuan-Anh and Nguyen-Truong, Hai and Ha, Tan-Sang and Pham, Quan-Dung and Yeung, Sai-Kit and Feng, Yuan and Thien, Nguyen Thanh and Tian, Lixin and Michel, Andreas and Gross, Wolfgang and Weinmann, Martin and Carrillo-Perez, Borja and Klein, Alexander and Alex, Antje and Solano-Carrillo, Edgardo and Steiniger, Yannik and Rodriguez, Angel Bueno and Kuan, Sheng-Yao and Ho, Yuan-Hao and Sattler, Felix and Fabijani\'c, Matej and \v{S}imunec, Magdalena and Kapetanovi\'c, Nadir}, title = {2nd Workshop on Maritime Computer Vision (MaCVi) 2024: Challenge Results}, booktitle = {Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) Workshops}, month = {January}, year = {2024}, pages = {869-891} }
2nd Workshop on Maritime Computer Vision (MaCVi) 2024: Challenge Results
Abstract
The 2nd Workshop on Maritime Computer Vision (MaCVi) 2024 continues to emphasize maritime computer vision for Unmanned Aerial Vehicles (UAV) and Unmanned Surface Vehicles (USV). Building on the foundation laid during the inaugural workshop, this year introduces several new challenges and datasets: (i) UAV-based Maritime Object Tracking with Reidentification using the SeaDronesSee-MOT dataset, (ii) USV-based Maritime Obstacle Segmentation and Detection with the new LaRS dataset, and (iii) USV-based Maritime Boat Tracking using the BoaTrack dataset. To address the lack of embedded methods, the Obstacle Segmentation features an embedded sub-track. This report offers a comprehensive overview of the findings from these challenges. We provide both statistical and qualitative analyses, evaluating trends from over 100 submissions. Detailed methodologies are summarized in the appendix. All datasets, evaluation codes, and the leaderboard are available to the public at https://macvi.org/workshop/ macvi24.
Related Material