RampWatch: An in-the-wild dataset and text-guided detection framework for recreational vessels
Abstract
Detecting small, recreational vessels in coastal environments remains a challenge due to complex backgrounds, dynamic lighting conditions, and the scarcity of annotated data for non-commercial maritime traffic. Despite their socio-economic significance, recreational boats are underrepresented in existing datasets and are poorly detected by standard object detectors, particularly in open-vocabulary scenarios. To address this gap, we present RampWatch, an in-the-wild dataset curated from surveillance footage at multiple boat ramps. RampWatch provides instance annotations across 7 categories of recreational vessels, captured under diverse weather, lighting, and occlusion conditions. To benchmark detection in this domain, we introduce YOLO-TG, a novel detection framework that augments YOLOv11 with a text encoder for open-vocabulary recognition and a self-attention module for enhanced spatial reasoning. YOLO-TG adopts a dual-stream design: visual features are extracted via a hierarchical YOLO backbone, while semantic embeddings from natural language prompts are encoded by a frozen language encoder. These modalities are fused via lightweight cross-modal attention, enabling text-guided detection without retraining. YOLO-TG achieves a +12% relative improvement in mAP@50-95 over strong YOLOv11 baselines on RampWatch, and demonstrates robust cross-domain generalization, with gains of +22% on the Singapore Maritime Dataset and +4.3% on the Split Port Ship Classification Dataset. These results highlight the effectiveness of cross-modal grounding and domain-specific datasets for advancing open-world maritime surveillance.
Keywords
computer vision, deep learning, maritime surveillance, zero-shot object detection
Document Type
Conference Proceeding
Date of Publication
1-1-2026
Publication Title
2026 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)
Publisher
IEEE
School
Centre for Artificial Intelligence and Machine Learning (CAIML) / School of Science
ISBN
[9798331555115]
Copyright
subscription content
Content Type
Metadata only
First Page
7576
Last Page
7585
Recommended Citation
Asim, M. M., Smallwood, C. B., Tariq, A., Lo, J., & Gilani, S. Z. (2026). RampWatch: An in-the-wild dataset and text-guided detection framework for recreational vessels. In Proceedings of the 2026 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) (pp. 7576-7585). IEEE. https://doi.org/10.1109/WACV61042.2026.00731