graviti
Products
Resources
About us
CURE-TSD
2D Box
Classification
Autonomous Driving
|...
License: MIT

Overview

We investigate the robustness of traffic sign recognition algorithms under challenging conditions.
Existing datasets are limited in terms of their size and challenging condition coverage, which
motivated us to generate the Challenging Unreal and Real Environments for Traffic Sign Recognition
(CURE-TSR) dataset. It includes more than two million traffic sign images that are based on
real-world and simulator data. We benchmark the performance of existing solutions in real-world
scenarios and analyze the performance variation with respect to challenging conditions. We
show that challenging conditions can decrease the performance of baseline methods significantly,
especially if these challenging conditions result in loss or misplacement of spatial information.
We also investigate the effect of data augmentation and show that utilization of simulator
data along with real-world data enhance the average recognition performance in real-world scenarios.

Data Format

File Name Format

“sequenceType_sequenceNumber_challengeSourceType_challengeType_challengeLevel.mp4”

  • sequenceType: 01 – Real data 02 – Unreal data
  • sequenceNumber: A number in between [01 – 49]
  • challengeSourceType: 00 – No challenge source (which means no challenge) 01 – After affect
  • challengeType: 00 – No challenge 01 – Decolorization 02 – Lens blur 03 – Codec error 04 –
    Darkening 05 – Dirty lens 06 – Exposure 07 – Gaussian blur 08 – Noise 09 – Rain 10 – Shadow
    11 – Snow 12 – Haze
  • challengeLevel: A number in between [01-05] where 01 is the least severe
    and 05 is the most severe challenge.

Test Sequences

We split the video sequences into 70% training set and
30% test set. The sequence numbers corresponding to test set are given below:

[01_04_x_x_x,
01_05_x_x_x, 01_06_x_x_x, 01_07_x_x_x, 01_08_x_x_x, 01_18_x_x_x, 01_19_x_x_x, 01_21_x_x_x,
01_24_x_x_x, 01_26_x_x_x, 01_31_x_x_x, 01_38_x_x_x, 01_39_x_x_x, 01_41_x_x_x, 01_47_x_x_x,
02_02_x_x_x, 02_04_x_x_x, 02_06_x_x_x, 02_09_x_x_x, 02_12_x_x_x, 02_13_x_x_x, 02_16_x_x_x,
02_17_x_x_x, 02_18_x_x_x, 02_20_x_x_x, 02_22_x_x_x, 02_28_x_x_x, 02_31_x_x_x, 02_32_x_x_x,
02_36_x_x_x]

The videos with all other sequence numbers are in the training set. Note that
“x” above refers to the variations listed earlier.

Coordinate System

img

Annotation Format

“sequenceType_sequenceNumber.txt“.

Challenge source type, challenge type, and challenge level do not affect the annotations. Therefore,
the video sequences that start with the same sequence type and the sequence number have the
same annotations.

  • sequenceType: 01 – Real data 02 – Unreal data
  • sequenceNumber: A number in between [01 – 49]

The format of each line in the annotation file (txt) should be: “frameNumber_signType_llx_lly_lrx_lry_ulx_uly_urx_ury”.

  • frameNumber: A number in between [001-300]
  • signType: 01 – speed_limit 02 – goods_vehicles 03 – no_overtaking 04 – no_stopping 05 – no_parking
    06 – stop 07 – bicycle 08 – hump 09 – no_left 10 – no_right 11 – priority_to 12 – no_entry
    13 – yield 14 – parking

Instruction

The name format of the video files are as follows:
“sequenceType_sequenceNumber_challengeSourceType_challengeType_challengeLevel.mp4”

  • sequenceType: 01 – Real data 02 – Unreal data
  • sequenceNumber: A number in between [01 – 49]
  • challengeSourceType: 00 – No challenge source (which means no challenge) 01 – After affect
  • challengeType: 00 – No challenge 01 – Decolorization 02 – Lens blur 03 – Codec error 04
    -Darkening 05 – Dirty lens 06 – Exposure 07 – Gaussian blur 08 – Noise 09 – Rain 10 – Shadow
    11 – Snow 12 – Haze * challengeLevel: A number in between [01-05] where 01 is the least severe
    and 05 is the most severe challenge.

Citation

If you use CURE-TSD dataset or codes, please cite the papers listed below:

Traffic Sign Detection Under Challenging Conditions: A Deeper Look into Performance Variations
and Spectral Characteristics

@ARTICLE{temel2019traffic,
author={D. Temel and M. Chen and G. AlRegib},
journal={IEEE Transactions on Intelligent Transportation Systems},
title={Traffic Sign Detection
Under Challenging Conditions: A Deeper Look into Performance Variations and Spectral Characteristics},
year={2019},
volume={},
number={},
pages={1-11},
doi={10.1109/TITS.2019.2931429},
ISSN={1524-9050},
url={https://arxiv.org/abs/1908.11262}}

[Traffic Signs in the Wild: Highlights from the IEEE Video and Image Processing Cup 2017 Student
Competition SP Competitions]

@ARTICLE{Temel2018_SPM,
author={D. Temel and G. AlRegib},
journal={IEEE Sig. Proc. Mag.},
title={Traffic Signs in
the Wild: Highlights from the IEEE Video and Image Processing Cup 2017 Student
Competition [SP Competitions]},
year={2018},
volume={35},
number={2},
pages={154-161},
doi={10.1109/MSP.2017.2783449},
ISSN={1053-5888},
url={https://arxiv.org/abs/1810.06169}}

Challenging
Environments for Traffic Sign Detection: Reliability Assessment under Inclement Conditions

@article{temel2019challenging,
  title={Challenging environments for traffic sign detection: Reliability assessment under
inclement conditions},
  author={Temel, Dogancan and Alshawi, Tariq and Chen, Min-Hung and AlRegib, Ghassan},
  journal={arXiv preprint arXiv:1902.06857},
  year={2019},
  url={https://arxiv.org/abs/1902.06857}
}

CURE-TSR: Challenging unreal and real environments for traffic sign
recognition

@INPROCEEDINGS{Temel2017_NIPSW,
Author = {D. Temel and G. Kwon and M. Prabhushankar and G. AlRegib},
Title = {{CURE-TSR: Challenging unreal and real environments for traffic sign recognition}},
Year = {2017},
booktitle = {Neural Information Processing
Systems (NeurIPS) Workshop on Machine Learning for Intelligent Transportation Systems},

License

MIT

Data Summary
Type
Video,
Amount
5.733K
Size
233.61GB
Provided by
OLIVES Lab
The Georgia Institute of Technology is a leading research university committed to improving the human condition through advanced science and technology.
| Amount 5.733K | Size 233.61GB
CURE-TSD
2D Box Classification
Autonomous Driving
License: MIT

Overview

We investigate the robustness of traffic sign recognition algorithms under challenging conditions.
Existing datasets are limited in terms of their size and challenging condition coverage, which
motivated us to generate the Challenging Unreal and Real Environments for Traffic Sign Recognition
(CURE-TSR) dataset. It includes more than two million traffic sign images that are based on
real-world and simulator data. We benchmark the performance of existing solutions in real-world
scenarios and analyze the performance variation with respect to challenging conditions. We
show that challenging conditions can decrease the performance of baseline methods significantly,
especially if these challenging conditions result in loss or misplacement of spatial information.
We also investigate the effect of data augmentation and show that utilization of simulator
data along with real-world data enhance the average recognition performance in real-world scenarios.

Data Format

File Name Format

“sequenceType_sequenceNumber_challengeSourceType_challengeType_challengeLevel.mp4”

  • sequenceType: 01 – Real data 02 – Unreal data
  • sequenceNumber: A number in between [01 – 49]
  • challengeSourceType: 00 – No challenge source (which means no challenge) 01 – After affect
  • challengeType: 00 – No challenge 01 – Decolorization 02 – Lens blur 03 – Codec error 04 –
    Darkening 05 – Dirty lens 06 – Exposure 07 – Gaussian blur 08 – Noise 09 – Rain 10 – Shadow
    11 – Snow 12 – Haze
  • challengeLevel: A number in between [01-05] where 01 is the least severe
    and 05 is the most severe challenge.

Test Sequences

We split the video sequences into 70% training set and
30% test set. The sequence numbers corresponding to test set are given below:

[01_04_x_x_x,
01_05_x_x_x, 01_06_x_x_x, 01_07_x_x_x, 01_08_x_x_x, 01_18_x_x_x, 01_19_x_x_x, 01_21_x_x_x,
01_24_x_x_x, 01_26_x_x_x, 01_31_x_x_x, 01_38_x_x_x, 01_39_x_x_x, 01_41_x_x_x, 01_47_x_x_x,
02_02_x_x_x, 02_04_x_x_x, 02_06_x_x_x, 02_09_x_x_x, 02_12_x_x_x, 02_13_x_x_x, 02_16_x_x_x,
02_17_x_x_x, 02_18_x_x_x, 02_20_x_x_x, 02_22_x_x_x, 02_28_x_x_x, 02_31_x_x_x, 02_32_x_x_x,
02_36_x_x_x]

The videos with all other sequence numbers are in the training set. Note that
“x” above refers to the variations listed earlier.

Coordinate System

img

Annotation Format

“sequenceType_sequenceNumber.txt“.

Challenge source type, challenge type, and challenge level do not affect the annotations. Therefore,
the video sequences that start with the same sequence type and the sequence number have the
same annotations.

  • sequenceType: 01 – Real data 02 – Unreal data
  • sequenceNumber: A number in between [01 – 49]

The format of each line in the annotation file (txt) should be: “frameNumber_signType_llx_lly_lrx_lry_ulx_uly_urx_ury”.

  • frameNumber: A number in between [001-300]
  • signType: 01 – speed_limit 02 – goods_vehicles 03 – no_overtaking 04 – no_stopping 05 – no_parking
    06 – stop 07 – bicycle 08 – hump 09 – no_left 10 – no_right 11 – priority_to 12 – no_entry
    13 – yield 14 – parking

Instruction

The name format of the video files are as follows:
“sequenceType_sequenceNumber_challengeSourceType_challengeType_challengeLevel.mp4”

  • sequenceType: 01 – Real data 02 – Unreal data
  • sequenceNumber: A number in between [01 – 49]
  • challengeSourceType: 00 – No challenge source (which means no challenge) 01 – After affect
  • challengeType: 00 – No challenge 01 – Decolorization 02 – Lens blur 03 – Codec error 04
    -Darkening 05 – Dirty lens 06 – Exposure 07 – Gaussian blur 08 – Noise 09 – Rain 10 – Shadow
    11 – Snow 12 – Haze * challengeLevel: A number in between [01-05] where 01 is the least severe
    and 05 is the most severe challenge.

Citation

If you use CURE-TSD dataset or codes, please cite the papers listed below:

Traffic Sign Detection Under Challenging Conditions: A Deeper Look into Performance Variations
and Spectral Characteristics

@ARTICLE{temel2019traffic,
author={D. Temel and M. Chen and G. AlRegib},
journal={IEEE Transactions on Intelligent Transportation Systems},
title={Traffic Sign Detection
Under Challenging Conditions: A Deeper Look into Performance Variations and Spectral Characteristics},
year={2019},
volume={},
number={},
pages={1-11},
doi={10.1109/TITS.2019.2931429},
ISSN={1524-9050},
url={https://arxiv.org/abs/1908.11262}}

[Traffic Signs in the Wild: Highlights from the IEEE Video and Image Processing Cup 2017 Student
Competition SP Competitions]

@ARTICLE{Temel2018_SPM,
author={D. Temel and G. AlRegib},
journal={IEEE Sig. Proc. Mag.},
title={Traffic Signs in
the Wild: Highlights from the IEEE Video and Image Processing Cup 2017 Student
Competition [SP Competitions]},
year={2018},
volume={35},
number={2},
pages={154-161},
doi={10.1109/MSP.2017.2783449},
ISSN={1053-5888},
url={https://arxiv.org/abs/1810.06169}}

Challenging
Environments for Traffic Sign Detection: Reliability Assessment under Inclement Conditions

@article{temel2019challenging,
  title={Challenging environments for traffic sign detection: Reliability assessment under
inclement conditions},
  author={Temel, Dogancan and Alshawi, Tariq and Chen, Min-Hung and AlRegib, Ghassan},
  journal={arXiv preprint arXiv:1902.06857},
  year={2019},
  url={https://arxiv.org/abs/1902.06857}
}

CURE-TSR: Challenging unreal and real environments for traffic sign
recognition

@INPROCEEDINGS{Temel2017_NIPSW,
Author = {D. Temel and G. Kwon and M. Prabhushankar and G. AlRegib},
Title = {{CURE-TSR: Challenging unreal and real environments for traffic sign recognition}},
Year = {2017},
booktitle = {Neural Information Processing
Systems (NeurIPS) Workshop on Machine Learning for Intelligent Transportation Systems},

License

MIT

0
Start building your AI now
graviti
wechat-QR
Long pressing the QR code to follow wechat official account

Copyright@Graviti