DSS-pytorch icon indicating copy to clipboard operation
DSS-pytorch copied to clipboard

:star: PyTorch implement of Deeply Supervised Salient Object Detection with Short Connection

Results 27 DSS-pytorch issues
Sort by recently updated
recently updated
newest added

As the author discribed in the chapter 4.5 in their TPAMI paper, they add another branch to do the classification for dataset including background images. But I am nor clear...

你好,感谢你所做的贡献 我想请问一下,我运行的我的数据集时候,会出现‘Can not pickle local object 'get_loader,.'’的错误,我在下载你们的MSRA数据集的时候,只有image和mask,并没生成train,test和val的txt,所以我想请问一下,这三个txt是如何生成的,内部格式如何?

Please help me. I want to save the all output results to "results/test". How can I do for this?

我进行训练和前向推理,发现需要测试图片相应的真值图,才能进行测试,那我要是自己的图片,我要如何进行前向推理(前向推理是要得到显著性图,而真值图我是没有的)

你好!慕名前来~几个问题,希望能得到街道 1、我看到你使用的backbone是vgg,通过extract_vgg.py下载得到的文件,大小是528M,其和你提供的vgg_feat不一致,请问有何区别,并且vgg_feat是如何得到的? 2、在问题1的基础上,如果我使用resnet作为backbone,我知道可以从网址下载得到其参数文件,但正如问题1,是否对应vgg_feat也有一个resnet_feat呢? 谢谢啦~

I wonder where is the --- test_cvpr2013.txt --- train_cvpr2013.txt --- valid_cvpr2013.txt --- test_cvpr2013_debug.txt --- train_cvpr2013_debug.txt --- valid_cvpr2013_debug.txt could you give me the link? thanks

I don't understand what's the means about this: extra = {'dss': .....} Could you tell me what's this and how to do with it.Thank you very much.

I find that there is a little distance between your results and the results in paper.Have you get a better result? I'am training this model now, can you give me...

In F-measure "score = (1 + self.beta ** 2) * avg_prec * avg_recall / (self.beta ** 2 * avg_prec + avg_recall)" and defined the self.beat=0.3 this is error