[visionlist] Call for Challenge: Hyperspectral City V1.0 with ICCV2019 workshop on Physics Based Vision meets Deep Learning

[Shaodi YOU]

=================================================================

CALL FOR CHALLENGE

HYPER SPECTRAL CITY V1.0 CHALLENGE

In Conjunct with The 2nd Workshop on Physics Based Vision Meets Deep
Learning (PBDL2019)

In Conjunction, November 2, Seoul, Korea

https://pbdl2019.github.io/

------------------------------------------------------------

We propose a semantic segmentation challenge for urban autonomous driving
scene which utilizes newly developed hyperspectral camera. The motivation
is to compensate the insufficient visual quality problem of existing
dataset. Particularly, the CityScape[1] dataset provides only extremely
washed out RGB images. To solve this, we endeavour to propose the new
dataset which adopts multi-channel visual input. Our new dataset, can
provide the following benefits: 1. properly balanced and colourful visual
input. 2. We can analyse and see visual properties which cannot be seen
from RGB channels. 3. We can robustly handle night scenes, thanks to the
near infrared band. 4. We can robustly handle water phenomenon including
rain and fog, because of the absolution behaviour in the infrared band.



For the initial release of the dataset, we decide to propose the task of
semantic segmentation with coarse labeling. We release 367 frames
hyperspectral images with coarse labeling for training and 67 frames with
fine labeling for testing.





------------------------------------------------------------

Important Dates:



    - Site open and training datset avaialbe for download: July 25,2019

    - Testing dataset avaialbe for download, challenge open for submission:
Aug 8,2019

    - Challenge close: Aug 28,2019

------------------------------------------------------------

Access:

The dataset access is through registration at:
http://pbdl2019.natapp1.cc/download

The Dataset is for this challenge and research purpose only. Commercial use
is not permitted.



If participants would like to use dataset in this challenge to produce
academic outputs, this document must be cited:



[1]"Hyperspectral City V1.0 Dataset and Benchmark”, Shaodi You*,
arXiv:1907.10270.



E-mail pbdl.ws at outlook.com if you have any technical problems.

------------------------------------------------------------

Organizers:

Yu Li, Tencent, China

Ying Fu, Beijing Institute of Technology, China

Shaodi You, Data61-CSIRO, Australia

Yinqiang Zheng, National Institute of Informatics, Japan

Feng Lu, Beihang University, China

Boxin Shi, Peking University, China

Robby T. Tan, Yale-NUS College & National University of Singapore, Singapore

------------------------------------------------------------

Contact:

For any inquiries regarding the workshop, please contact: pbdl.ws at gmail.com

E-mail pbdl.ws at outlook.com if you have any technical problems.

=================================================================


-- 
YOU, Shaodi
youshaodi at gmail.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://visionscience.com/pipermail/visionlist_visionscience.com/attachments/20190725/b3654a8b/attachment.html>