jdai-cv / veridataset Goto Github PK

Hi,

can you provide please the intrinsic and extrinsic parameters from the cameras used to extract the images?

great dataset by the way :-)

thanks in advance

I am studying the subject of pedestrian re-identification and I sincerely ask you for a data set--VeRi dataset.

In your busy schedule, I sent you an email.

Looking forward to your reply, thank you very much！

I sent an email and stated my name and organization. The dataset will be not for commercial use, but only for the completion of the experiment. But no reply has been received yet.

I think the SOTA evaluation metrics in README should be mAP, Rank-1, Rank-5, instead of current Rank-1, Rank-5, mAP.

没有数据集下载地址啊，联系邮箱也没用了

How can I submit my reidentification results to publish them on the leaderboard?

The label file for VeRI train_label.xml version 1.0 (<TrainingImages Version="1.0">) has the following typeID errors:

Vehicle 3, 7, and 41 in the training set are labeled as typeID="0" which does not correspond to a legal type in list_type.txt.

Vehicle 0003 is an Audi Q5, which we believe is typeID="2" (SUV).
Vehicle 0007 is a Nissan of some sort. It looks like a cross-over, so again typeID="2" (SUV).
Vehicle 0041 is a Hummer H2, which should be typeID="2" (SUV).

I don’t find any paper or references about the method “ours”.
But the rank1 and rank5 of the method "ours" are really good.
Is there any references of this method？
Thank you.

Hi,I‘m a postgraduate of Beijing Institute of technology and recently start to learn the field of traffic flow analysis.And I have read your codes,one question bothered me: you have the line “distance = int(input("Enter the length of the selected region in meters: "))”in your main.py, but I don't understand what is the "distance" stand for? In my understanding,your codes ask the user to input a quadrilateral at first(formed by four points ),Is the "quadrilateral" means "selected region"?So what the length of the "selected region"(distance) stand for?
Looking forward to your replay.

the password of the baidudisk about VeRi,can you tell me?

@VehicleReId
Thanks for your work!
My name is QiaoPeng, and I'm a student from Xidian University. I've been working on vehicle detection and track. Recently I'm trying to use deep_sort algorithm to track cars, since the deep_sort is aimed for track of pedestrian. I need to train deep_sort model with cars dataset to improve the performance of deep_sort algorithm. Then I found this dataset, I really need this dataset for my research. I couldn't find your email address, so I write this reply, hopefully you can see it and help me get the dataset. here are my email address [email protected]. looking forward to hear from you soon. Thankyou so much !

We have a newly published ICCV 2019 paper with experiments on VeRi-776. Hope it could be added to the state-of-the-art results and references.

@inproceedings{Tang19PAMTRI,
author = {Zheng Tang and Milind Naphade and Stan Birchfield and Jonathan Tremblay and William Hodge and Ratnesh Kumar and Shuo Wang and Xiaodong Yang},
title = {{PAMTRI}: {P}ose-aware multi-task learning for vehicle re-identification using highly randomized synthetic data},
booktitle = {Proc. ICCV},
pages = {211--220},
address = {Seoul, Korea},
year = {2019}
}

Link: http://openaccess.thecvf.com/content_ICCV_2019/papers/Tang_PAMTRI_Pose-Aware_Multi-Task_Learning_for_Vehicle_Re-Identification_Using_Highly_Randomized_ICCV_2019_paper.pdf

Our state-of-the-art performance is as follow (presented in Table 2 of the paper):
mAP: 71.88%
Rank-1: 92.86%
Rank-5: 96.97%

Can someone help me with understanding the ground truth format please? I can't understand how to write code to calculate the accuracy of my model.