Acknowledge license (cc-by-nc-nd-4.0) to accept the repository
This repository is publicly accessible, but you have to accept the conditions to access its files and content.
The access request is only accepted after we receive your signed form via email. Our team may take 2-3 days to process your request.
Log in or Sign Up to review the conditions and access this dataset content.
TSBOW: Traffic Surveillance Benchmark for Occluded Vehicles
Under Various Weather Conditions
Contact for Dataset: [email protected], [email protected], [email protected]
(UPDATING....)
(All links would be updated on the conference day.)
🎉 NEWS
- [2025.11.16] 🔥 Our code and website are released!
- [2025.11.08] 🎉 TSBOW has been accepted to AAAI 2026!
🌍 Overview
Comprehensive, annotated dataset for object detection. This dataset consists of over 32 hours of real-world traffic data across 71 CCTV and 1 additional color cameras, spanning annual weather conditions.
Abstract
Global warming has intensified the frequency and severity of extreme weather events, which degrade CCTV signal and video quality while disrupting traffic flow, thereby increasing traffic accident rates. Existing datasets, often limited to light haze, rain, and snow, fail to capture extreme weather conditions. To address this gap, this study introduces the Traffic Surveillance Benchmark for Occluded Vehicles under Various Weather Conditions (TSBOW), a comprehensive dataset designed to enhance occluded vehicle detection across diverse annual weather scenarios. Comprising over 32 hours of real-world traffic data from densely populated urban areas, TSBOW includes more than 48,000 manually annotated and 3.2 million semi-labeled frames; bounding boxes spanning eight traffic participant classes from large vehicles to micromobility devices and pedestrians. We establish an object detection benchmark for TSBOW, highlighting challenges posed by occlusions and adverse weather. With its varied road types, scales, and viewpoints, TSBOW serves as a critical resource for advancing Intelligent Transportation Systems. Our findings underscore the potential of CCTV-based traffic monitoring, paving the way for new research and applications. The TSBOW dataset is publicly available at the following link. **Code** -- https://github.com/SKKUAutoLab/TSBOW
198📹 Processed Videos 📹 |
32 h⏱️ Duration ⏱️ |
3.2 M🖼️ Total Frames 🖼️ |
71.1 MSemi-Annotated |
48 KManual-Annotated |
1.1 MManual-Anotated |
📊 Statistics
Recording Locations
Video Distribution
📥 Dataset Download
The TSBOW dataset is distributed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. By completing the form below, users acknowledge and agree that the dataset will be used solely for research purposes.
Submission Guidelines
1. TSBOW - Terms and Conditions Form
- Download and fill out the TSBOW - Terms and Conditions form.
- Ensure all User Information fields are completed.
- Provide a description of your intended use of the dataset.
- Check the agreement box and insert your handwritten signature.
- Save as PDF file. Renaming as: {TSBOW}{Application-Date}{Huggingface-Username}.pdf (i.e. TSBOW_20260120_ngochdm.pdf)
2. Requirement Email
- The subject format: [TSBOW Access Requirement] {Your Name} - {Affiliation}
- The email body includes your HuggingFace account information (username and email). We will verify this information against the access requirements on Hugging Face before approval.
- Send email to all our addresses: [email protected], [email protected], [email protected]
3. Send Request on HuggingFace
- Press "Agree and send request to access TSBOW" button on HuggingFace. Our team may take 2-3 days to process your request.
Scripts to download TSBOW from HuggingFace are provided in utils folder. Please refer to the download_TSBOW.py for more details.
📁 Format
Video Format
- Video Standard: mp4 (H.265)
- Video Resolution: 1280 x 720
- Video Frame Rate: varied from 20 to 30 FPS
- Videos are named as
<video_id>.mp4
Image Format
- Image Standard: jpg
- Image Resolution: 1280 x 720
- Images extracted from videos are named as
<video_id>_<frame_id>.jpg.<frame_id>is 6 characters length.
Ground Truth Format
Annotations are provided in YOLO format with one *.txt file per image. If there are no objects in an image, no *.txt file is required. Each line in text file are the bounding box information of an object:
<class_id> <x_center> <y_center> <width> <height>
where
<class_id>(int): class index starting from 0x_center,y_center(double): box coordinates normalized xywh format (from 0 to 1)width,height(double): box width, height
Metadata Format
The CSV file provides metadata for each videos, including: scenario, daytime, weather, scale, roadtype, video_id, total_duration, and ROI zone.
SCENARIO(char): one of four values:r(road),i(intersection),s(special cases),d(disaster).DAYTIME(char):d(day),n(night).WEATHER(char): one of four values:n(normal),h(haze),r(rain),s(snow).SCALE(char): one of three values:f(fine),m(medium),c(coarse).ROADTYPE(char): one of three values:u(urban),s(standard),b(boulevard).VIDEO_ID(string): video name.DURATION(duration): total duration of whole videos (test + val + train).ROI(string): a list of points in polygon.
Directory Structure
Train/Val/Test_Public:
videos.zip: video files.annotations.zip: annotations include images and labels.semilabels.zip: semi-labeled annotations include labels only. You can extract frames from videos.- Notes for Test_Public: In the first publication, we only public annotations of the last 30% of test set. The remaining files will be public in the future.
Others:
comparison.zip: annotations of datasets' comparison in Experiment section, include images and labels.TSBOW_info.csv: metadata file.
🏅 Citation
If our research is helpful to you, please cite our paper using the following BibTeX format
@article{Huynh_TSBOW_AAAI_2026,
title = {TSBOW: Traffic Surveillance Benchmark for Occluded Vehicles Under Various Weather Conditions},
author = {Ngoc Doan-Minh Huynh, Duong Nguyen-Ngoc Tran, Long Hoang Pham, Tai Huu-Phuong Tran, Hyung-Joon Jeon, Huy-Hung Nguyen, Duong Khac Vu, Hyung-Min Jeon, Son Hong Phan, Quoc Pham-Nam Ho, Chi Dai Tran, Trinh Le Ba Khanh, Jae Wook Jeon},
journal = {AAAI 2026},
year = {2025}
}
📝 Changelog
- Downloads last month
- 273