GithubHelp home page GithubHelp logo

apfnet's People

Contributors

yangmengmeng1997 avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar

apfnet's Issues

模型测试问题(pytorch版本)

直接下载RGBT234_ALL_Transformer.pth权重,在tracking/Run.py里对GTOT数据集测试,报错
_pickle.UnpicklingError: A load persistent id instruction was encountered,
but no persistent_load function was specified.
换了多个torch版本仍然未解决,代码运行卡在了权重加载部分。

请问您用的具体哪个python和torch版本
屏幕截图 2022-09-23 221113

屏幕截图 2022-09-23 221732

imagenet权重

您好,我看您的代码里面没有mdnet_imagenet_vid.pth权重,可以提供一下吗?谢谢

如何选择最终的模型权重呢

作者您好!请教问下您在最后一个阶段训练时,怎么选择最终的模型权重呢?我这边复现了您的实验,最后一阶段训练结束后,控制台输出的Mean Precision只有60%多,请问这个复现数据有问题吗?您当时达到多少了呢?

log文件问题

这个项目第二、三阶段的训练以及测试都需要log文件,请问项目中为什么没有log文件呀?

Challenge Tag

I'm trying to train APFNEt but I have a problem with creating .pkl files for different challenge types. There is no .tag file (for example fast_motion.tag) in RGBT-234 or GTOT datasets. Have you created these tag files on your own? Do you have end can you share them with me? If you don't have them, can you tell me how to create them? Thank you very much.

if challenge_type == 'FM': challenge_inf = 'fast_motion.tag' elif challenge_type == 'OCC': challenge_inf = 'occlusion.tag' elif challenge_type == 'SC': challenge_inf = 'size_change.tag' elif challenge_type == 'ILL': challenge_inf = 'illum_change.tag' elif challenge_type == 'TC': challenge_inf = 'thermal_crossover.tag' elif challenge_type == 'ALL': data[seqname] = {'images_v':img_list_v, 'images_i':img_list_i, 'gt':gt}

about CMPP results

hello!
Thank you very much for your contribution in the field of RGBT object tracking. I reproduced your code and got excellent results. I have a question, how did you obtain the results of CMPP and CAT during the evaluation, I did not find relevant data or codes. I would be very grateful if you could help me!

数据集的问题

您好,我想问一下您提供的RGBT234的数据集里面没有rgbt.txt文件,我在官方文件里面下载txt文件没问题吧?为什么一直提示我找不到文件呢?希望作者能解答一下,谢谢

Pretrained Model Link

Link of pretrained model is broken or I can't reach it because of Baidu's policies. Can you renew it or upload to another cloud platform except Baidu?

属性标签

作者您好,请问您知道lasher的数据集怎么生成或制作tag属性标签吗?我看GTOT和RGBT234数据集都有

下载的GTOT数据集里没有challenge tag

您好,这是一篇非常棒的工作!但我在复现的时候遇到一点小问题,我下载了GTOT数据集,但里面的目录里找不到所谓的challenge tag,下载数据集如下所示:
2022-05-12 20-05-44屏幕截图
这是我的数据集设置:
2022-05-12 20-09-33屏幕截图

请问这些challenge tag是需要预处理才能获得吗?

关于RGBT234数据集

您好!我非常喜欢您的工作,然而我也同样遇到了数据集的问题,带有init.txt标注文件的RGBT234数据集无法找到,项目中的百度云链接目前已经无法下载,请问能否发布新的下载链接?

关于预训练权重mdnet_imagenet_vid.pth的问题

作者你好,我在您的网络结构基础上更改了一些结构,这样就导致之前的mdnet_imagenet_vid.pth的初始化参数匹配不上,我想问如果我不采用mdnet_imagenet_vid.pth初始化权重直接手动初始化,会很难收敛吗

model_stage3代码问题

    output = torch.bmm(affinity, w_v) 
    # output=output.permute(0,2,1)
    # output=self.transformer1_FFN[1](nn.Dropout(0.2)(F.relu(((self.transformer1_FFN[0](output))))))
    # output=output.permute(0,2,1)
    output=output.reshape(batch,dim,w,h)
    output=self.transformer1_encoder1[3](output)  
    x=x+output   
    return x

你好,我想问一下按照您论文里面figure3的encoder的部分是先进行了自注意力后再add再升通道的,但是按照您model_stage3.py里面就是上面的代码是进行了自注意力后先升通道再与原始值相加的,请问是我理解有问题吗?

关于评测工具的问题

作者你好,为什么我直接从GitHub上直接下载下来的RGBT234的数据结果,用MATLAB跑上面说Precision是0.815,Success是0.561啊?我都没有修改

复现结果

您好,我想问一下在RGBT234上论文里的结果是82.7%/57.9%,GTOT的结果是90.5%/73.7%,我复现的结果在RGBT234上是80.4%/55.5%,GTOT上是88.4%/71.8%。直接使用您的模型在GTOT上跑的结果是88.7%/72.2%。请问这样的结果是正常的吗?我要怎样才能达到论文的精度?

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.