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

About the training data sampling? #6

Open ff521 opened 2 years ago

ff521 commented 2 years ago

作者您好!非常感谢您的开源代码! 请问一下譬如IP数据集训练数据的采样为何是如此分布的呢,它与其它HSI分类方法的采样方式有什么优势呢? 感谢您的分享!

danfenghong commented 2 years ago

你好,我们采用的是固定的训练和测试样本,具体如何选择的,是人工选择出具有挑战的样本,这种方式有助于公平的评估不同的算法,有利于我们重新产生结果。 祝好 丹枫

ff521 @.***> 于2022年1月4日周二 17:34写道:

作者您好!非常感谢您的开源代码! 请问一下譬如IP数据集训练数据的采样为何是如此分布的呢,它与其它HSI分类方法的采样方式有什么优势呢? 感谢您的分享!

— Reply to this email directly, view it on GitHub https://github.com/danfenghong/IEEE_TGRS_SpectralFormer/issues/6, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFL2GZTU5NDNFSAIXBNBCTTUUK5LNANCNFSM5LHAEOHA . 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: @.***>

ff521 commented 2 years ago

好的作者,我这里还有一个问题。 您的自注意力机制方法是应用在频谱上的,将光谱通道作为token输入到encoder中,也就是说transformer提取的是一个像元不同波段之间的相关性。 那么这个单像元光谱是一个一维向量,其中的相关性对于像元分类起到什么作用呢,因为还涉及到像元混叠等因素。空间域的相关性能够体现结构信息所以是有用的。 感谢您的回复!

------------------ 原始邮件 ------------------ 发件人: "danfenghong/IEEE_TGRS_SpectralFormer" @.>; 发送时间: 2022年1月4日(星期二) 晚上6:09 @.>; @.**@.>; 主题: Re: [danfenghong/IEEE_TGRS_SpectralFormer] About the training data sampling? (Issue #6)

你好,我们采用的是固定的训练和测试样本,具体如何选择的,是人工选择出具有挑战的样本,这种方式有助于公平的评估不同的算法,有利于我们重新产生结果。 祝好 丹枫

ff521 @.***> 于2022年1月4日周二 17:34写道:

> 作者您好!非常感谢您的开源代码! > 请问一下譬如IP数据集训练数据的采样为何是如此分布的呢,它与其它HSI分类方法的采样方式有什么优势呢? > 感谢您的分享! > > — > Reply to this email directly, view it on GitHub > <https://github.com/danfenghong/IEEE_TGRS_SpectralFormer/issues/6&gt;, or > unsubscribe > <https://github.com/notifications/unsubscribe-auth/AFL2GZTU5NDNFSAIXBNBCTTUUK5LNANCNFSM5LHAEOHA&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 are subscribed to this thread.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 authored the thread.Message ID: @.***>

danfenghong commented 2 years ago

看您的问题应该是做方法不是做遥感机理的。当然我的解释也不一定对。我的理解是,高的光谱分辨率能够通过一些更精细的变化区分出一些相似度比较高的地物,例如两种地物的光谱曲线很相似,但由于其物理特性,在这两种地物仅在某一些波段的反射吸收是不同的,这个时候光谱局部的信息就十分重要了。当然你说的光谱mixing或是结构之类的这些都是反应空间信息了,这个我们的论文也有使用patch作为输入的。其实这些说明和解释在论文也都有。 希望有所帮助! 祝好! Danfeng

ff521 @.***> 于2022年1月10日周一 11:47写道:

好的作者,我这里还有一个问题。 您的自注意力机制方法是应用在频谱上的,将光谱通道作为token输入到encoder中,也就是说transformer提取的是一个像元不同波段之间的相关性。

那么这个单像元光谱是一个一维向量,其中的相关性对于像元分类起到什么作用呢,因为还涉及到像元混叠等因素。空间域的相关性能够体现结构信息所以是有用的。 感谢您的回复!

------------------ 原始邮件 ------------------ 发件人: "danfenghong/IEEE_TGRS_SpectralFormer" @.>; 发送时间: 2022年1月4日(星期二) 晚上6:09 @.>; @.**@.>; 主题: Re: [danfenghong/IEEE_TGRS_SpectralFormer] About the training data sampling? (Issue #6)

你好,我们采用的是固定的训练和测试样本,具体如何选择的,是人工选择出具有挑战的样本,这种方式有助于公平的评估不同的算法,有利于我们重新产生结果。 祝好 丹枫

ff521 @.***> 于2022年1月4日周二 17:34写道:

> 作者您好!非常感谢您的开源代码! > 请问一下譬如IP数据集训练数据的采样为何是如此分布的呢,它与其它HSI分类方法的采样方式有什么优势呢? > 感谢您的分享! > > — > Reply to this email directly, view it on GitHub > <https://github.com/danfenghong/IEEE_TGRS_SpectralFormer/issues/6&gt;, or > unsubscribe > < https://github.com/notifications/unsubscribe-auth/AFL2GZTU5NDNFSAIXBNBCTTUUK5LNANCNFSM5LHAEOHA&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 are subscribed to this thread.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 authored the thread.Message ID: @.***>

— Reply to this email directly, view it on GitHub https://github.com/danfenghong/IEEE_TGRS_SpectralFormer/issues/6#issuecomment-1008520418, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFL2GZSKULASJ44NF7AVZ5LUVJJD3ANCNFSM5LHAEOHA . 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: @.***>