Explanation of the .txt annotations

[MuhammadAsadJaved]

Thank you for this explanation. It's very helpful. I have a few more questions about the dataset.

1 - There are 11 sets in total with several subsets.
Some .txt annotation contain only " % bbGt version=3" this tag and no bounding box values. What does this mean? Should we remove these annotations?

2- Can you explain the values in the .txt file? for example
annotation set00/V000/I02165 contains these values.

% bbGt version=3
person 427 243 27 66 0 0 0 0 0 0 0

T

he first one is class, next 4 are bounding box coordinates ,what about other values? Do we need these in the training ?

3- How we divide these annotation to 9 categories? Reasonable all, Reasonable day, Reasonable night, Near scale, Far scare, Medium scare, No occlusion, Partial occlusion, and Heavy occlusion? Anyone write any scripts to categories images and .txt files?
Please also see attached images for Question 1 and 2.

Thank you very much.

[diciembre-noche]

Hello Asad, I'm a master student currently studying object detection. I would really like to try on this dataset for research purpose.
Could you please share the dataset with me since the download link is dead? Full set or some subsets are all fine.
I'm really appreciated to any of your help. And perhaps I can help you solve your doubts after having a look of the data. My email is lovepasta@gmx.net
Thank you!!

[MuhammadAsadJaved]

OK. I'll get back to you soon.

…

On Tue, Aug 11, 2020 at 6:52 PM diciembre-noche ***@***.***> wrote: Hello Asad, I'm a master student currently studying object detection. I would really like to try on this dataset for research purpose. Could you please share the dataset with me since the download link is dead? Full set or some subsets are all fine. I'm really appreciated to any of your help. And perhaps I can help you solve your doubts after having a look of the data. My email is ***@***.*** Thank you!! — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#18 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AG4GR5GAT44IO6P2PIOOJ7DSAEPFBANCNFSM4PDJSIBQ> .

[MuhammadAsadJaved]

@diciembre-noche

Can you please try this link , if you do not understand the distribution of dataset then let me know I'll explain to you

https://soonminhwang.github.io/rgbt-ped-detection/data/

[diciembre-noche]

@MuhammadAsadJaved
Thanks for Sharing! Unfortunately your page seems to be not working either. It has the same dead links from this https://sites.google.com/site/pedestrianbenchmark/home
Do you download from this page recently?
Is it convenient for you to share any subset from day and night? For example set01 and set04.

[MuhammadAsadJaved]

@MuhammadAsadJaved
Thanks for Sharing! Unfortunately your page seems to be not working either. It has the same dead links from this https://sites.google.com/site/pedestrianbenchmark/home
Do you download from this page recently?
Is it convenient for you to share any subset from day and night? For example set01 and set04.

@diciembre-noche

Sorry, Use this link.

You can download complete sets or download sets in parts. Do not download
both. For example, download set00.zip (if you have a fast internet
connection to download big files)
or alternatively download set00_V000.zip, set00_V001.zip, set00_V00n
(small sets for slow internet connection). Don't download both,

https://onedrive.live.com/?authkey=%21ADG6wuQeYqCroBI&id=1570430EADF56512%21624&cid=1570430EADF56512

if you failed to use above link, then use this link given below to submit a request and they will send you a link in the given email.

https://sites.google.com/site/pedestrianbenchmark/download

[diciembre-noche]

@MuhammadAsadJaved Thank you so much! I actually did request via that link and got the mail days ago. But I somehow miss that link because most of the subset links are dead....so careless..
Anyway thanks again for your help and hope you everything good with your work:))

[MuhammadAsadJaved]

Brother I downloaded using the same link, dataset is up to date. There might be problem in your internet. If you still can’t download then let me know i will upload required set for you.

…

On Wed, Aug 12, 2020 at 4:41 PM diciembre-noche ***@***.***> wrote: @MuhammadAsadJaved <https://github.com/MuhammadAsadJaved> Thank you so much! I actually did request via that link and got the mail days ago. But I somehow miss that link because most of the subset links are dead....so careless.. Anyway thanks again for your help and hope you everything good with your work:)) — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#18 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AG4GR5EIANGW4VAC5UIHKG3SAJISRANCNFSM4PDJSIBQ> .

[fnsflm]

You can see it,
https://pdollar.github.io/toolbox/
And click detector->bbGt.

Each object struct has the following fields:
lbl - a string label describing object type (eg: 'pedestrian')
bb - [l t w h]: bb indicating predicted object extent
occ - 0/1 value indicating if bb is occluded
bbv - [l t w h]: bb indicating visible region (may be [0 0 0 0])
ign - 0/1 value indicating bb was marked as ignore
ang - [0-360] orientation of bb in degrees