search

danfenghong / IEEE_TGRS_SpectralFormer

Danfeng Hong, Zhu Han, Jing Yao, Lianru Gao, Bing Zhang, Antonio Plaza, Jocelyn Chanussot. Spectralformer: Rethinking hyperspectral image classification with transformers, IEEE Transactions on Geoscience and Remote Sensing (TGRS), 2021
219 stars 39 forks source link

What's the mean of the 'TE' and 'TR' in your '.mat' dataset? #7

Open kangpeilun opened 2 years ago

kangpeilun commented 2 years ago

I search the google and the baidu, but I still dont konw the mean of 'TE' and 'TR'. Your code use them to make up the label. If I want to user my own dataset, what should I do to generate my '.mat' data like yours?

danfenghong commented 2 years ago

Ok, I see. About the TR and TE, it is also separated by experts. If you would like to use your own data, you can simply randomly select some points for training and the rest for testing.

Cheers, Danfeng

东北林业大学 @.***> 于2022年1月5日周三 21:12写道：

I search the google and the baidu, but I still dont konw the mean of 'TE' and 'TR'. Your code use them to make up the label. If I want to user my own dataset, what should I do to generate my '.mat' data like yours?

— Reply to this email directly, view it on GitHub https://github.com/danfenghong/IEEE_TGRS_SpectralFormer/issues/7, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFL2GZRKNKGVB4HVJU5WK6DUUQ7UFANCNFSM5LJ353LQ . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you are subscribed to this thread.Message ID: @.***>

kangpeilun commented 2 years ago

oh, I see. Thanks

liuzuo-byte commented 2 years ago

作者你好，我研一小硕，知识储备不够，最近在看你的论文，我想问论文里面的分组操作，x这个像素点有m个波段，我看结构图，是不是3个波段为一组呢？分成m组？每组3个波段，那按照公式，第一个组，就是x0，x1,x2,这3个波段为1组？那没有x0啊，怎么弄啊？

danfenghong commented 2 years ago

你好，大致思想是这样的，图像是画的3个一组，具体选几个这是一个参数需要选择的，我们的后面实验也有讨论。如果X0没有你可以做镜像或是复制操作，就像是做卷积边缘做镜像是一个道理。

liuzuo-byte @.***> 于2022年1月6日周四 11:36写道：

作者你好，我研一小硕，知识储备不够，最近在看你的论文，我想问论文里面的分组操作，x这个像素点有m个波段，我看结构图，是不是3个波段为一组呢？分成m组？每组3个波段，那按照公式，第一个组，就是x0，x1,x2,这3个波段为1组？那没有x0啊，怎么弄啊？

— Reply to this email directly, view it on GitHub https://github.com/danfenghong/IEEE_TGRS_SpectralFormer/issues/7#issuecomment-1006258847, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFL2GZRG4ZF7TBS36RUEAHLUUUE2RANCNFSM5LJ353LQ . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you commented.Message ID: @.***>

liuzuo-byte commented 2 years ago

谢谢回复，论文里面的64 unit是指选64个波段为一组吗，77指的是选49个像素点？77大小的cube？

------------------ 原始邮件 ------------------ 发件人: "danfenghong/IEEE_TGRS_SpectralFormer" @.>; 发送时间: 2022年1月6日(星期四) 中午12:35 @.>; @.**@.>; 主题: Re: [danfenghong/IEEE_TGRS_SpectralFormer] What's the mean of the 'TE' and 'TR' in your '.mat' dataset? (Issue #7)

你好，大致思想是这样的，图像是画的3个一组，具体选几个这是一个参数需要选择的，我们的后面实验也有讨论。如果X0没有你可以做镜像或是复制操作，就像是做卷积边缘做镜像是一个道理。

liuzuo-byte @.***> 于2022年1月6日周四 11:36写道：

> > 作者你好，我研一小硕，知识储备不够，最近在看你的论文，我想问论文里面的分组操作，x这个像素点有m个波段，我看结构图，是不是3个波段为一组呢？分成m组？每组3个波段，那按照公式，第一个组，就是x0，x1,x2,这3个波段为1组？那没有x0啊，怎么弄啊？ > > — > Reply to this email directly, view it on GitHub > <https://github.com/danfenghong/IEEE_TGRS_SpectralFormer/issues/7#issuecomment-1006258847&gt;, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AFL2GZRG4ZF7TBS36RUEAHLUUUE2RANCNFSM5LJ353LQ&gt; > . > Triage notifications on the go with GitHub Mobile for iOS > <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&amp;mt=8&amp;pt=524675&gt; > or Android > <https://play.google.com/store/apps/details?id=com.github.android&amp;referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub&gt;. > > You are receiving this because you commented.Message ID: > @.***> >

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you commented.Message ID: @.***>

liuzuo-byte commented 2 years ago

我不明白这里的n代表考虑了n个像素点还是n波段为一组

------------------ 原始邮件 ------------------ 发件人: "danfenghong/IEEE_TGRS_SpectralFormer" @.>; 发送时间: 2022年1月6日(星期四) 中午12:35 @.>; @.**@.>; 主题: Re: [danfenghong/IEEE_TGRS_SpectralFormer] What's the mean of the 'TE' and 'TR' in your '.mat' dataset? (Issue #7)

你好，大致思想是这样的，图像是画的3个一组，具体选几个这是一个参数需要选择的，我们的后面实验也有讨论。如果X0没有你可以做镜像或是复制操作，就像是做卷积边缘做镜像是一个道理。

liuzuo-byte @.***> 于2022年1月6日周四 11:36写道：

> > 作者你好，我研一小硕，知识储备不够，最近在看你的论文，我想问论文里面的分组操作，x这个像素点有m个波段，我看结构图，是不是3个波段为一组呢？分成m组？每组3个波段，那按照公式，第一个组，就是x0，x1,x2,这3个波段为1组？那没有x0啊，怎么弄啊？ > > — > Reply to this email directly, view it on GitHub > <https://github.com/danfenghong/IEEE_TGRS_SpectralFormer/issues/7#issuecomment-1006258847&gt;, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AFL2GZRG4ZF7TBS36RUEAHLUUUE2RANCNFSM5LJ353LQ&gt; > . > Triage notifications on the go with GitHub Mobile for iOS > <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&amp;mt=8&amp;pt=524675&gt; > or Android > <https://play.google.com/store/apps/details?id=com.github.android&amp;referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub&gt;. > > You are receiving this because you commented.Message ID: > @.***> >

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you commented.Message ID: @.***>

danfenghong commented 2 years ago

64unit是指网络的神经元数量

liuzuo-byte @.***> 于2022年1月6日周四 13:01写道：

谢谢回复，论文里面的64 unit是指选64个波段为一组吗，77指的是选49个像素点？77大小的cube？

------------------ 原始邮件 ------------------ 发件人: "danfenghong/IEEE_TGRS_SpectralFormer" @.>; 发送时间: 2022年1月6日(星期四) 中午12:35 @.>; @.**@.>; 主题: Re: [danfenghong/IEEE_TGRS_SpectralFormer] What's the mean of the 'TE' and 'TR' in your '.mat' dataset? (Issue #7)

你好，大致思想是这样的，图像是画的3个一组，具体选几个这是一个参数需要选择的，我们的后面实验也有讨论。如果X0没有你可以做镜像或是复制操作，就像是做卷积边缘做镜像是一个道理。

liuzuo-byte @.***> 于2022年1月6日周四 11:36写道：

> > 作者你好，我研一小硕，知识储备不够，最近在看你的论文，我想问论文里面的分组操作，x这个像素点有m个波段，我看结构图，是不是3个波段为一组呢？分成m组？每组3个波段，那按照公式，第一个组，就是x0，x1,x2,这3个波段为1组？那没有x0啊，怎么弄啊？

> > — > Reply to this email directly, view it on GitHub > < https://github.com/danfenghong/IEEE_TGRS_SpectralFormer/issues/7#issuecomment-1006258847&gt;,

> or unsubscribe > < https://github.com/notifications/unsubscribe-auth/AFL2GZRG4ZF7TBS36RUEAHLUUUE2RANCNFSM5LJ353LQ&gt;

> . > Triage notifications on the go with GitHub Mobile for iOS > < https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&amp;mt=8&amp;pt=524675&gt;

> or Android > < https://play.google.com/store/apps/details?id=com.github.android&amp;referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub&gt;.

> > You are receiving this because you commented.Message ID: > @.***> >

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you commented.Message ID: @.***>

— Reply to this email directly, view it on GitHub https://github.com/danfenghong/IEEE_TGRS_SpectralFormer/issues/7#issuecomment-1006286272, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFL2GZXNUDYGJNLFZAE4DCDUUUOZTANCNFSM5LJ353LQ . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you commented.Message ID: @.***>

liuzuo-byte commented 2 years ago

谢谢回复，我看了好几个礼拜没懂，感谢作者回复。还一个问题就是用patch为77的cube为输入的时候，X1,….Xm 为每个波段展开，那么我选3个波段为一组的时候，这时3个波段组成为493的二维矩阵，用上面GSE模块的W做embading的时候，得到的是一个二维矩阵呀？不是一个向量。

发自我的iPhone

------------------ 原始邮件 ------------------ 发件人: Danfeng Hong @.> 发送时间: 2022年1月6日 14:29 收件人: danfenghong/IEEE_TGRS_SpectralFormer @.> 抄送: liuzuo-byte @.>, Comment @.> 主题: Re: [danfenghong/IEEE_TGRS_SpectralFormer] What's the mean of the 'TE' and 'TR' in your '.mat' dataset? (Issue #7)

64unit是指网络的神经元数量

liuzuo-byte @.***> 于2022年1月6日周四 13:01写道：

> 谢谢回复，论文里面的64 unit是指选64个波段为一组吗，77指的是选49个像素点？77大小的cube？ > > > > > ------------------&nbsp;原始邮件&nbsp;------------------ > 发件人: "danfenghong/IEEE_TGRS_SpectralFormer" @.&gt;; > 发送时间:&nbsp;2022年1月6日(星期四) 中午12:35 > @.&gt;; > @.**@.&gt;; > 主题:&nbsp;Re: [danfenghong/IEEE_TGRS_SpectralFormer] What's the mean of the > 'TE' and 'TR' in your '.mat' dataset? (Issue #7) > > > > > > 你好，大致思想是这样的，图像是画的3个一组，具体选几个这是一个参数需要选择的，我们的后面实验也有讨论。如果X0没有你可以做镜像或是复制操作，就像是做卷积边缘做镜像是一个道理。 > > > liuzuo-byte @.&gt; 于2022年1月6日周四 11:36写道： > > &gt; > &gt; > 作者你好，我研一小硕，知识储备不够，最近在看你的论文，我想问论文里面的分组操作，x这个像素点有m个波段，我看结构图，是不是3个波段为一组呢？分成m组？每组3个波段，那按照公式，第一个组，就是x0，x1,x2,这3个波段为1组？那没有x0啊，怎么弄啊？ > > &gt; > &gt; — > &gt; Reply to this email directly, view it on GitHub > &gt; < > https://github.com/danfenghong/IEEE_TGRS_SpectralFormer/issues/7#issuecomment-1006258847&amp;gt;, > > &gt; or unsubscribe > &gt; < > https://github.com/notifications/unsubscribe-auth/AFL2GZRG4ZF7TBS36RUEAHLUUUE2RANCNFSM5LJ353LQ&amp;gt; > > &gt; . > &gt; Triage notifications on the go with GitHub Mobile for iOS > &gt; < > https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&amp;amp;mt=8&amp;amp;pt=524675&amp;gt; > > &gt; or Android > &gt; < > https://play.google.com/store/apps/details?id=com.github.android&amp;amp;referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub&amp;gt;. > > &gt; > &gt; You are receiving this because you commented.Message ID: > &gt; @.&gt; > &gt; > > — > Reply to this email directly, view it on GitHub, or unsubscribe. > Triage notifications on the go with GitHub Mobile for iOS or Android. > You are receiving this because you commented.Message ID: @.&gt; > > — > Reply to this email directly, view it on GitHub > <https://github.com/danfenghong/IEEE_TGRS_SpectralFormer/issues/7#issuecomment-1006286272&gt;, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AFL2GZXNUDYGJNLFZAE4DCDUUUOZTANCNFSM5LJ353LQ&gt; > . > Triage notifications on the go with GitHub Mobile for iOS > <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&amp;mt=8&amp;pt=524675&gt; > or Android > <https://play.google.com/store/apps/details?id=com.github.android&amp;referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub&gt;. > > You are receiving this because you commented.Message ID: > @.> >

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you commented.Message ID: @.***>

danfenghong commented 2 years ago

是要拉成向量的，不是矩阵，代码和论文都能看出来。

liuzuo-byte @.***> 于2022年1月6日周四 14:39写道：

谢谢回复，我看了好几个礼拜没懂，感谢作者回复。还一个问题就是用patch为77的cube为输入的时候，X1,….Xm 为每个波段展开，那么我选3个波段为一组的时候，这时3个波段组成为493的二维矩阵，用上面GSE模块的W做embading的时候，得到的是一个二维矩阵呀？不是一个向量。

发自我的iPhone

------------------ 原始邮件 ------------------ 发件人: Danfeng Hong @.> 发送时间: 2022年1月6日 14:29 收件人: danfenghong/IEEE_TGRS_SpectralFormer @.> 抄送: liuzuo-byte @.>, Comment @.> 主题: Re: [danfenghong/IEEE_TGRS_SpectralFormer] What's the mean of the 'TE' and 'TR' in your '.mat' dataset? (Issue #7)

64unit是指网络的神经元数量

liuzuo-byte @.***> 于2022年1月6日周四 13:01写道：

> 谢谢回复，论文里面的64 unit是指选64个波段为一组吗，77指的是选49个像素点？77大小的cube？ > > > > > ------------------&nbsp;原始邮件&nbsp;------------------ > 发件人: "danfenghong/IEEE_TGRS_SpectralFormer" @.&gt;; > 发送时间:&nbsp;2022年1月6日(星期四) 中午12:35 > @.&gt;; > @.**@.&gt;; > 主题:&nbsp;Re: [danfenghong/IEEE_TGRS_SpectralFormer] What's the mean of the > 'TE' and 'TR' in your '.mat' dataset? (Issue #7) > > > > > > 你好，大致思想是这样的，图像是画的3个一组，具体选几个这是一个参数需要选择的，我们的后面实验也有讨论。如果X0没有你可以做镜像或是复制操作，就像是做卷积边缘做镜像是一个道理。

> > > liuzuo-byte @.***&gt; 于2022年1月6日周四 11:36写道： > > &gt; > &gt; > 作者你好，我研一小硕，知识储备不够，最近在看你的论文，我想问论文里面的分组操作，x这个像素点有m个波段，我看结构图，是不是3个波段为一组呢？分成m组？每组3个波段，那按照公式，第一个组，就是x0，x1,x2,这3个波段为1组？那没有x0啊，怎么弄啊？

> > &gt; > &gt; — > &gt; Reply to this email directly, view it on GitHub > &gt; < > https://github.com/danfenghong/IEEE_TGRS_SpectralFormer/issues/7#issuecomment-1006258847&amp;gt;,

> > &gt; or unsubscribe > &gt; < > https://github.com/notifications/unsubscribe-auth/AFL2GZRG4ZF7TBS36RUEAHLUUUE2RANCNFSM5LJ353LQ&amp;gt;

> > &gt; . > &gt; Triage notifications on the go with GitHub Mobile for iOS > &gt; < > https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&amp;amp;mt=8&amp;amp;pt=524675&amp;gt;

> > &gt; or Android > &gt; < > https://play.google.com/store/apps/details?id=com.github.android&amp;amp;referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub&amp;gt;.

> > &gt; > &gt; You are receiving this because you commented.Message ID: > &gt; @.&gt; > &gt; > > — > Reply to this email directly, view it on GitHub, or unsubscribe. > Triage notifications on the go with GitHub Mobile for iOS or Android. > You are receiving this because you commented.Message ID: @.&gt; > > — > Reply to this email directly, view it on GitHub > < https://github.com/danfenghong/IEEE_TGRS_SpectralFormer/issues/7#issuecomment-1006286272&gt;,

> or unsubscribe > < https://github.com/notifications/unsubscribe-auth/AFL2GZXNUDYGJNLFZAE4DCDUUUOZTANCNFSM5LJ353LQ&gt;

> . > Triage notifications on the go with GitHub Mobile for iOS > < https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&amp;mt=8&amp;pt=524675&gt;

> or Android > < https://play.google.com/store/apps/details?id=com.github.android&amp;referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub&gt;.

> > You are receiving this because you commented.Message ID: > @.***> >

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you commented.Message ID: @.***>

— Reply to this email directly, view it on GitHub https://github.com/danfenghong/IEEE_TGRS_SpectralFormer/issues/7#issuecomment-1006321696, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFL2GZTGAYG3CVWX5MKB54TUUU2JRANCNFSM5LJ353LQ . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you commented.Message ID: @.***>

liuzuo-byte commented 2 years ago

谢谢🙏，祝作者身体健康，工作顺利！

发自我的iPhone

------------------ 原始邮件 ------------------ 发件人: Danfeng Hong @.> 发送时间: 2022年1月6日 16:50 收件人: danfenghong/IEEE_TGRS_SpectralFormer @.> 抄送: liuzuo-byte @.>, Comment @.> 主题: Re: [danfenghong/IEEE_TGRS_SpectralFormer] What's the mean of the'TE' and 'TR' in your '.mat' dataset? (Issue #7)

danfenghong commented 2 years ago

多谢对我们工作的兴趣和支持。

liuzuo-byte @.***> 于2022年1月6日周四 16:54写道：

谢谢🙏，祝作者身体健康，工作顺利！

发自我的iPhone

------------------ 原始邮件 ------------------ 发件人: Danfeng Hong @.> 发送时间: 2022年1月6日 16:50 收件人: danfenghong/IEEE_TGRS_SpectralFormer @.> 抄送: liuzuo-byte @.>, Comment @.> 主题: Re: [danfenghong/IEEE_TGRS_SpectralFormer] What's the mean of the'TE' and 'TR' in your '.mat' dataset? (Issue #7)

— Reply to this email directly, view it on GitHub https://github.com/danfenghong/IEEE_TGRS_SpectralFormer/issues/7#issuecomment-1006387162, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFL2GZVVREQYI5RB3EXCMITUUVKFHANCNFSM5LJ353LQ . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you commented.Message ID: @.***>