layumi / AICIty-reID-2020

:red_car: The 1st Place Submission to AICity Challenge 2020 re-id track (Baidu-UTS submission)
MIT License
455 stars 108 forks source link

测试所得mAP很低。 #18

Open s20163081161 opened 4 years ago

s20163081161 commented 4 years ago

@郑博士 你好。我这边运行train+virtual数据集,当运行到40个Epoch的时候,太慢了,我把它停了,然后跑测试集,但结果rank@5才不到90%,mAP才0.35,这正常吗??会是哪里出错了呢??

layumi commented 4 years ago

 60 和 75 轮会降learning rate https://github.com/layumi/AICIty-reID-2020/blob/master/pytorch/train_2020.py#L540

应该是正常的

s20163081161 commented 4 years ago

60 和 75 轮会降learning rate https://github.com/layumi/AICIty-reID-2020/blob/master/pytorch/train_2020.py#L540 应该是正常的

但都已经训练了30多个Epoch了,rank@5还不到90%,这也正常吗??

Kang9779 commented 4 years ago

@郑博士 你好。我这边运行train+virtual数据集,当运行到40个Epoch的时候,太慢了,我把它停了,然后跑测试集,但结果rank@5才不到90%,mAP才0.35,这正常吗??会是哪里出错了呢??

你有这个比赛的数据集吗?

s20163081161 commented 4 years ago

@郑博士 你好。我这边运行train+virtual数据集,当运行到40个Epoch的时候,太慢了,我把它停了,然后跑测试集,但结果rank@5才不到90%,mAP才0.35,这正常吗??会是哪里出错了呢??

你有这个比赛的数据集吗?

网上申请即可

s20163081161 commented 4 years ago

@layumi 郑博士,我用真实数据集(也就是400个目标)进行训练,把80个Epoch都跑完了,但是mAP仍是只有0.47,请问这正常吗?

layumi commented 4 years ago

@s20163081161 是的。差不多。因为我validation用的gallery比较大。 之后你可以尝试post processing之类的,继续提升。

s20163081161 commented 4 years ago

@s20163081161 是的。差不多。因为我validation用的gallery比较大。 之后你可以尝试post processing之类的,继续提升。

好的,多谢。

s20163081161 commented 4 years ago

@s20163081161 是的。差不多。因为我validation用的gallery比较大。 之后你可以尝试post processing之类的,继续提升。

关于您所说的“gallery比较大”,我看您这里边的gallery是36900多张图片,那么多大的数量算大呢??有没有个衡量便准啥的??

HanwenCao commented 4 years ago

@郑博士 你好。我这边运行train+virtual数据集,当运行到40个Epoch的时候,太慢了,我把它停了,然后跑测试集,但结果rank@5才不到90%,mAP才0.35,这正常吗??会是哪里出错了呢??

您好@s20163081161,请问您跑的测试集是怎么制作的?我看readme只讲了训练集的制作。谢谢!

s20163081161 commented 4 years ago

网上申请即可

---原始邮件--- 发件人: "Hanwen Cao"<notifications@github.com> 发送时间: 2020年6月28日(周日) 下午5:57 收件人: "layumi/AICIty-reID-2020"<AICIty-reID-2020@noreply.github.com>; 抄送: "Mention"<mention@noreply.github.com>;"s20163081161"<742915081@qq.com>; 主题: Re: [layumi/AICIty-reID-2020] 测试所得mAP很低。 (#18)

@郑博士 你好。我这边运行train+virtual数据集,当运行到40个Epoch的时候,太慢了,我把它停了,然后跑测试集,但结果rank@5才不到90%,mAP才0.35,这正常吗??会是哪里出错了呢??

您好@s20163081161,请问您跑的测试集是怎么制作的?我看readme只讲了训练集的制作。谢谢!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

HanwenCao commented 4 years ago

您好@s20163081161,打扰了,我在网上申请并得到了所有数据集。您说的mAP的测试代码是submit.py还是test_2020.py?

s20163081161 commented 4 years ago

test_2020.py。你跑的如何?

---原始邮件--- 发件人: "Hanwen Cao"<notifications@github.com> 发送时间: 2020年6月28日(周日) 晚上6:12 收件人: "layumi/AICIty-reID-2020"<AICIty-reID-2020@noreply.github.com>; 抄送: "Mention"<mention@noreply.github.com>;"s20163081161"<742915081@qq.com>; 主题: Re: [layumi/AICIty-reID-2020] 测试所得mAP很低。 (#18)

您好@s20163081161,打扰了,我在网上申请并得到了所有数据集。您说的mAP的测试代码是submit.py还是test_2020.py?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

HanwenCao commented 4 years ago

test_2020.py。你跑的如何? ---原始邮件--- 发件人: "Hanwen Cao"<notifications@github.com> 发送时间: 2020年6月28日(周日) 晚上6:12 收件人: "layumi/AICIty-reID-2020"<AICIty-reID-2020@noreply.github.com>; 抄送: "Mention"<mention@noreply.github.com>;"s20163081161"<742915081@qq.com>; 主题: Re: [layumi/AICIty-reID-2020] 测试所得mAP很低。 (#18) 您好@s20163081161,打扰了,我在网上申请并得到了所有数据集。您说的mAP的测试代码是submit.py还是test_2020.py? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

抱歉,还没跑成功 :/

s20163081161 commented 4 years ago

哈哈,跑成功了说一声

---原始邮件--- 发件人: "Hanwen Cao"<notifications@github.com> 发送时间: 2020年6月28日(周日) 晚上6:36 收件人: "layumi/AICIty-reID-2020"<AICIty-reID-2020@noreply.github.com>; 抄送: "Mention"<mention@noreply.github.com>;"s20163081161"<742915081@qq.com>; 主题: Re: [layumi/AICIty-reID-2020] 测试所得mAP很低。 (#18)

test_2020.py。你跑的如何? … ---原始邮件--- 发件人: "Hanwen Cao"<notifications@github.com> 发送时间: 2020年6月28日(周日) 晚上6:12 收件人: "layumi/AICIty-reID-2020"<AICIty-reID-2020@noreply.github.com>; 抄送: "Mention"<mention@noreply.github.com>;"s20163081161"<742915081@qq.com>; 主题: Re: [layumi/AICIty-reID-2020] 测试所得mAP很低。 (#18) 您好@s20163081161,打扰了,我在网上申请并得到了所有数据集。您说的mAP的测试代码是submit.py还是test_2020.py? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

抱歉,还没跑成功 :/

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

vokhidovhusan commented 3 years ago

@layumi @s20163081161 I am running test_2020.py and as @layumi has mentioned earlier I am going to use val for test. But how could I use val as a test dataset? Should i split it to gallery and query? Should I use val dataset as gallery or both gallery and query?

layumi commented 3 years ago

@s20163081161 test_2020 is to verify the effectiveness of the model on validation ( no offical valid set, so I splitted ). The final test result is obtained by the submission to the official test website.

layumi commented 3 years ago

@martianvenusian Test set is provided by the offical aicity organizer. If you want to verify the final result, please run submit_result.py. But they do not provide valid set, so I splitted the set to verify the model by using test_2020.py

vokhidovhusan commented 3 years ago

@layumi Thank you for your support. I have split val set to gallery and query as following if it is right way:

val/000341/ -- 000341_c010_7.jpg -- 000341_c016_11.jpg -- 000341_c017_5.jpg

query/000341/ -- 000341_c010_7.jpg

gallery/000341/ -- 000341_c016_11.jpg -- 000341_c017_5.jpg

hongrunqv commented 3 years ago

@郑博士 你好。我这边运行train+virtual数据集,当运行到40个Epoch的时候,太慢了,我把它停了,然后跑测试集,但结果rank@5才不到90%,mAP才0.35,这正常吗??会是哪里出错了呢??

你好,我想问一下在prepare_2020.py中生成的gallery这个文件夹,prepare_2020.py并没有向里面添加图片,但是在后面大量用到,这个文件夹是用来做什么的,谢谢

wanglaotou commented 3 years ago

@layumi Thank you for your support. I have split val set to gallery and query as following if it is right way:

val/000341/ -- 000341_c010_7.jpg -- 000341_c016_11.jpg -- 000341_c017_5.jpg

query/000341/ -- 000341_c010_7.jpg

gallery/000341/ -- 000341_c016_11.jpg -- 000341_c017_5.jpg

hello, could you please tell how to split the vaild set, which has query and gallery...

xh7682 commented 3 years ago

@郑博士 你好。我这边运行train+virtual数据集,当运行到40个Epoch的时候,太慢了,我把它停了,然后跑测试集,但结果rank@5才不到90%,mAP才0.35,这正常吗??会是哪里出错了呢??

您好@s20163081161,请问您跑的测试集是怎么制作的?我看readme只讲了训练集的制作。谢谢! 您好,请问gallery这个文件夹的数据是如何生成的呢?