File "run.py", line 61, in main l_data, u_data = get_init_shot_in_cam1(dataset_all, load_path="./examples/{}_init_{}.pickle".format(dataset_all.name, args.init), init=args.init) AttributeError: 'Market1501' object has no attribute 'name' about one-example-person-reid

[Yu-Wu]

Please check the content in ``examples''.

[ly-jy]

File "run.py", line 61, in main
l_data, u_data = get_init_shot_in_cam1(dataset_all, load_path="./examples/market1501_init_0.2.pickle".format(dataset_all.name, args.init), init=args.init)
AttributeError: 'Market1501' object has no attribute 'name'

[Yu-Wu]

@ly-jy
Please use original code.

There is attribute 'name' for this object.

One-Example-Person-ReID/reid/datasets/market1501.py

Line 12 in 829fb58

self.name="market1501"

[ly-jy]

When I run "sh run.sh" in the terminal,：
Files already downloaded and verified
Market1501 dataset loaded
subset | # ids | # images

train | 651 | 11329
val | 100 | 1607
trainval | 751 | 12936
query | 750 | 16483
gallery | 751 | 19281
Traceback (most recent call last):
File "run.py", line 125, in
main(parser.parse_args())
File "run.py", line 61, in main
l_data, u_data = get_init_shot_in_cam1(dataset_all, load_path="./examples/{}init{}.pickle".format(dataset_all.name, args.init), init=args.init)
AttributeError: 'Market1501' object has no attribute 'name'
How can I use the original code you said?

[ly-jy]

Excuse me, are you sure the code can run?
Can you run to make sure it can run?
I tried my best and didn't run through it.
Please help me, thank you very much!

[Yu-Wu]

@ly-jy

Please run this code under a clean python environment.
It seems you might have installed'' the reid module of other codes before. Therefore, the module Market1501'' might be referred to a previous installed path.

[ly-jy]

Thank you for your patience and answer. I hope you can help me. My problems are as follows:

I downloaded the code to the 'home' and upzip it. Put in the data set and run "sh run.sh" in the terminal. Report the error:
Traceback (most recent call last):
File "run.py", line 2, in
from reid.eug import *
ModuleNotFoundError: No module named 'reid.eug'

I cut out the 'eug.py' and modify ‘ from reid.eug import * ’ to ' ‘from eug import * ', And run 'sh run.sh'.Report the error:
Traceback (most recent call last):
File "run.py", line 2, in
from eug import *
File "/home/ly/liyang/One-Example-Person-ReID-master/eug.py", line 17, in
from reid.exclusive_loss import ExLoss
ModuleNotFoundError: No module named 'reid.exclusive_loss'

I cut out the 'exclusive_loss.py' and modify ‘ from reid.exclusive_loss import ExLoss * ’ to ' from exclusive_loss import ExLoss * ', And run 'sh run.sh'.Report the error:
Files already downloaded and verified
Market1501 dataset loaded
subset | # ids | # images

train | 651 | 11228
val | 100 | 1708
trainval | 751 | 12936
query | 750 | 16483
gallery | 751 | 19281
Traceback (most recent call last):
File "run.py", line 125, in
main(parser.parse_args())
File "run.py", line 61, in main
l_data, u_data = get_init_shot_in_cam1(dataset_all, load_path="./examples/{}init{}.pickle".format(dataset_all.name, args.init), init=args.init)
AttributeError: 'Market1501' object has no attribute 'name'

I change ' l_data, u_data = get_init_shot_in_cam1(dataset_all, load_path="./examples/{}init{}.pickle".format(dataset_all.name, args.init), init=args.init) ' to ' l_data, u_data = get_init_shot_in_cam1(dataset_all, load_path="./examples/market1501_init_0.2.pickle".format(dataset_all.name, args.init), init=args.init) '. But still an error：
File "run.py", line 61, in main
l_data, u_data = get_init_shot_in_cam1(dataset_all, load_path="./examples/market1501_init_0.2.pickle".format(dataset_all.name, args.init), init=args.init)
AttributeError: 'Market1501' object has no attribute 'name'

How can I do?

[Yu-Wu]

@ly-jy

You should not fix the issue in this way......

I have updated my code just now. You can try the new code.
Please make sure you are using python3.

[ly-jy]

Thank you very much for your answer. I have used python3, but I have encountered the following problem:
I downloaded the code to the 'home' and upzip it. I run "sh run.sh" in the terminal. Report the error:
FileNotFoundError: [Errno 2] No such file or directory: '/home/ly/One-Example-Person-ReID-master/data/market1501/train

I changed the ‘bounding_box_train’ file in the market to ‘train’, 'bounding_box_test' to 'gallery'.Then I put ‘train’, 'gallery', 'query' into the '/home/ly/One-Example-Person-ReID-master/data/market1501' file. And run 'sh run.sh'. Report the error:
NotADirectoryError: [Errno 20] Not a directory: '/home/ly/One-Example-Person-ReID-master/data/market1501/train/0002_c1s1_000451_03.jpg'
File '0002_c1s1_000451_03.jpg' already exists in file 'train'.How can I do?

[Yu-Wu]

@ly-jy
You should use my changed dataset since I changed file names to make sure Image dataset work the same as the video dataset.

Please download them using the provided dataset link in readme.

[ly-jy]

Hello author, I have some thoughts after reading your article. I want to know how the datasets you provide are processed. Can you share your code that processes the dataset? I am very grateful! Plase help me! Thank you very much!

[ly-jy]

@ly-jy
You should use my changed dataset since I changed file names to make sure Image dataset work the same as the video dataset.

Please download them using the provided dataset link in readme.

I want to know how the datasets you provide are changed. Can you share your code that change the dataset? I am very grateful! Plase help me! Thank you very much!

[Yu-Wu]

Sorry that the processing code is lost.

For your own dataset, you can use a very simple python scripts to generate the same format as ours. To do this, you only need some basic file operations.

[Yu-Wu]

You can see the difference between the original Market dataset and my provided one. It's very simple to generate our format.

You can also see https://github.com/Yu-Wu/DukeMTMC-VideoReID/blob/master/README.md to better understand my "Directory Structure"

[ly-jy]

Thank you very much for your answer. If you are convenient, I hope you can provide the code. Thank you so much！Plase help me! Thank you very much!

[Yu-Wu]

Sorry that the code is lost.

I cannot find it.