Weather

Videos for the TSBOW dataset were recorded in Suwon under diverse weather condition throughout the year: normal, haze, rain, and snow.

• Normal: sunny and cloudy
• Haze: haze or foggy
• Rain: raindrops or active rainfall are evident
• Snow: visible snowflkes or snowfall

Scenario

The scenarios are categorized into four distinct types: road, intersection, special cases, and disaster.

• Road: straight roads or those where traffi flow remains unaffected by traffi lights
• Intersection: includes three subtypes—T, Y, or crossroad—and extends to scenes where traffi lights or pedestrian crossings inflence traffi flow
• Special cases: includes videos featuring shared lanes, overpasses, or mid-road construction
• Disaster: pertains to scenarios where hostile weather severely degrades video quality

Road Type

The TSBOW dataset classifies data collection zones by the number of straight lanes per direction, excluding turning lanes, into three road types: urban, standard, and boulevard.

• Urban: two lanes
• Standard: four lanes
• Boulevard: over six lanes

Scale

CCTV cameras along routes and intersections vary in angle and height, producing bounding box sizes categorized into three scales: fine, medium, and coarse.

• Fine: near road surface, detailed object visualization
• Medium: elevated cameras, discernible license plates but reduced clarity
• Coarse: distant cameras, high vehicle count but partial visibility due to distance and occlusion

Traffic

Traffic flow is categorized into light, moderate, and heavy, based on the maximum number of bounding boxes per frame in video.

• Light: less than 25 bboxes/frame
• Moderate: 25 to less than 50 bboxes/frame
• Heavy: over 50 bboxes/frame

Cite Our Work

If our research is helpful to you, please cite our paper
using the following BibTeX format

Citation will be updated again after AAAI-26 proceedings publication.

@misc{huynh2026tsbowtrafficsurveillancebenchmark,
      title={TSBOW: Traffic Surveillance Benchmark for Occluded Vehicles Under Various Weather Conditions}, 
      author={Ngoc Doan-Minh Huynh and Duong Nguyen-Ngoc Tran and Long Hoang Pham and Tai Huu-Phuong Tran and Hyung-Joon Jeon and Huy-Hung Nguyen and Duong Khac Vu and Hyung-Min Jeon and Son Hong Phan and Quoc Pham-Nam Ho and Chi Dai Tran and Trinh Le Ba Khanh and Jae Wook Jeon},
      year={2026},
      eprint={2602.05414},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2602.05414}, 
}

📋 Copy