Zhedong Zheng comments

Results 584 comments of


                                            Zhedong Zheng

Create a custom Market-1501 dataset

Hi @Jaroslove Please follow the folder setting in the tutorial. The main idea is to keep one ID for one subfolder.

dataset labels

@ps-star Yes. Please check [[Latitude and Longitude]](https://drive.google.com/file/d/1PL8fVky9KZg7XESsuS5NCsYRyYAwui3S/view?usp=sharing) Thanks a lot. I just highlighted the info in the first several lines of README.md.

使用 fuse_all_conv_bn 之后推理速度更慢

Thanks for reporting! I have not tested it. I will take care.

请求邮件拒收

Thank you for reporting. I just sent it.

How to generate top-10 answer.txt file

Thank you. You may refer to the previous competition code. https://github.com/wtyhub/UAVM2024 like https://github.com/koorter/WELN/blob/main/export.ipynb

请求邮件被拒收

@chenqi142 Sorry. I just saw the issue. The link has been sent to your email. Please take a look.

test.py

Are you running `test.py`? Please check https://github.com/layumi/University1652-Baseline/blob/master/test.py#L229 Did you modify the code? There are no any `for` or `while` loop functions.

test.py

Hi @Mayuzhuo0323 @ckmessi @hominsu Thank you @ps-star If you use Windows, please make sure your numw_workers=0. This is a well-kown case for pytorch on Windows. @ps-star Could you provide more...

About the dataset preparation

@8acemak1rkey 1. Google earth pro 2. Yes. I found the pattern and editted the kml file. Given the first location, you could generate the spiral flight.

我用的是你去年的版本，今年重新装了一个pytorch 训练densenet的时候正常但是测试的时候卡在第一个batchsize 不动了其他模型都没问题

@ZezeTao 你好你用的是windows还是ubuntu？可能是由于torch.compile 加速导致的。可以试试把对应的行注释掉。另外，densenet 再等一等能跑起来么？