Levigty / AimCLR

This is an official PyTorch implementation of "Contrastive Learning from Extremely Augmented Skeleton Sequences for Self-supervised Action Recognition" in AAAI2022.
MIT License
63 stars 14 forks source link

How to implement t-sne clustering #12

Open chaowen1 opened 1 year ago

chaowen1 commented 1 year ago

The cluster diagram in your readme is very attractive. Is there a corresponding part of the code that can draw this class diagram

yanghuaigang666 commented 1 year ago

Yes, I hope to share the source code!Thank you very much

Levigty commented 1 year ago

Thanks for your interest in our work. Regarding t-sne visualization, we refer to this project. You only need to save the features and labels (do not use but need to save) of the network on the pretext task as .pkl file, select some categories (like 10) and follow the readme of the project to get the visualization results.

chaowen1 commented 1 year ago

Thank you for your reply and wish you a happy life

沈超文… @.***

 

------------------ 原始邮件 ------------------ 发件人: "Levigty/AimCLR" @.>; 发送时间: 2023年2月20日(星期一) 下午2:56 @.>; @.**@.>; 主题: Re: [Levigty/AimCLR] How to implement t-sne clustering (Issue #12)

Thanks for your interest in our work. Regarding t-sne visualization, we refer to this project. You only need to save the features and labels (do not use but need to save) of the network on the pretext task as .pkl file, select some categories (like 10) and follow the readme of the project to get the visualization results.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

chaowen1 commented 1 year ago

Hello, I have saved the features and tags, but the code part of t-sne will not be implemented. Can I send a t-sne code

沈超文… @.***

 

------------------ 原始邮件 ------------------ 发件人: "Levigty/AimCLR" @.>; 发送时间: 2023年2月20日(星期一) 下午2:56 @.>; @.**@.>; 主题: Re: [Levigty/AimCLR] How to implement t-sne clustering (Issue #12)

Thanks for your interest in our work. Regarding t-sne visualization, we refer to this project. You only need to save the features and labels (do not use but need to save) of the network on the pretext task as .pkl file, select some categories (like 10) and follow the readme of the project to get the visualization results.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

Morris-Lucifer commented 1 year ago

Thanks for your interest in our work. Regarding t-sne visualization, we refer to this project. You only need to save the features and labels (do not use but need to save) of the network on the pretext task as .pkl file, select some categories (like 10) and follow the readme of the project to get the visualization results.

您好,请问对于文中的3s-AimCLR来说,保存用于t-sne的特征,是指选择每个stream的Normal Query Encoder编码后的特征进行保存,然后按照[0.6; 0.6; 0.4]加权融合后再进行可视化吗?

xievito commented 11 months ago

Hello, I have saved the features and tags, but the code part of t-sne will not be implemented. Can I send a t-sne code 沈超文… @.   ------------------ 原始邮件 ------------------ 发件人: "Levigty/AimCLR" @.>; 发送时间: 2023年2月20日(星期一) 下午2:56 @.>; @*.**@*.>; 主题: Re: [Levigty/AimCLR] How to implement t-sne clustering (Issue #12) Thanks for your interest in our work. Regarding t-sne visualization, we refer to this project. You only need to save the features and labels (do not use but need to save) of the network on the pretext task as .pkl file, select some categories (like 10) and follow the readme of the project to get the visualization results. — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.>

Hello, how do you save features and labels

panlongji commented 9 months ago

Does anyone have a solution?

zhouzhou1005 commented 3 weeks ago

Hello, I am very interested in your project and have tried to visualize it, but it failed. Could you please send me a copy of your visualization code?

zhouzhou1005 commented 3 weeks ago

Hello, I have saved the features and tags, but the code part of t-sne will not be implemented. Can I send a t-sne code 沈超文… @.   ------------------ 原始邮件 ------------------ 发件人: "Levigty/AimCLR" @.>; 发送时间: 2023年2月20日(星期一) 下午2:56 @.>; @*.**@*.>; 主题: Re: [Levigty/AimCLR] How to implement t-sne clustering (Issue #12) Thanks for your interest in our work. Regarding t-sne visualization, we refer to this project. You only need to save the features and labels (do not use but need to save) of the network on the pretext task as .pkl file, select some categories (like 10) and follow the readme of the project to get the visualization results. — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.>

Hello, I have saved the features and tags, but the code part of t-sne will not be implemented. Can I send a t-sne code 沈超文… @.   ------------------ 原始邮件 ------------------ 发件人: "Levigty/AimCLR" @.>; 发送时间: 2023年2月20日(星期一) 下午2:56 @.>; @*.**@*.>; 主题: Re: [Levigty/AimCLR] How to implement t-sne clustering (Issue #12) Thanks for your interest in our work. Regarding t-sne visualization, we refer to this project. You only need to save the features and labels (do not use but need to save) of the network on the pretext task as .pkl file, select some categories (like 10) and follow the readme of the project to get the visualization results. — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.>

May I ask how to save the features and labels? Thank you.