yangze0930 NTS-Net issues

yangze0930 / NTS-Net

This is a PyTorch implementation of the ECCV2018 paper "Learning to Navigate for Fine-grained Classification" (Ze Yang, Tiange Luo, Dong Wang, Zhiqiang Hu, Jun Gao, Liwei Wang).

MIT License

452 stars 118 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

能分享下权重吗，给的链接失效了~~

#57 atnegam opened 1 month ago
0
有人试过nts-net在小数据集的性能吗，我自定义数据集上面模型一直欠拟合underfitting，怎么回事呢

#56 cowangusm opened 1 year ago
0
test loss: 0.737 and test acc: 0.849 total sample: 5794

#55 durong1996 opened 2 years ago
0
utils and dataset implemented

#54 8kta opened 3 years ago
0
ValueError: not enough values to unpack (expected 2, got 0)

#53 bamboo821 opened 3 years ago
2
what's the aim of padding 0? what's the meaning of cat_num?

#52 newbieTo opened 3 years ago
2
为什么要在输入的图像四周补0，这样不会导致截取的区域全0么

#51 newbieTo closed 3 years ago
0
NTS-Net for small image size

#50 kitu-datta opened 3 years ago
0
模型文件无法下载？？

#49 banxianerr opened 3 years ago
0
Video nts net

#48 Linwei94 closed 2 years ago
0
error in resnet

#47 doaa1adel closed 3 years ago
0
Different test accuracy when testing with different BATCH_SIZE in config.py

#46 yangxh11 closed 4 years ago
5
This code version is sooo ~ bad! and memory error why???

#45 ballade8 opened 4 years ago
0
When I trained NTS-Net on my own dataset and stanford dog dataset, the train accuracy can reach 99.8%, but the test accuracy just 85%. Do you same as me

#44 Coderhancode opened 4 years ago
1
对于代码中ProposalNet有些疑问

#43 zhangyongbigdata opened 4 years ago
0
permissionError : [Errno 13 ] Permission denied : '/data_4t'

#42 zyk-bot opened 4 years ago
1
Set the rationality of the partcls_loss

#41 zel-1 opened 4 years ago
1
small batch_size, better result !

#40 huanyuxinchen opened 4 years ago
0
依赖库的版本问题

#39 hackjus opened 4 years ago
0
My owndataset has a SyntaxError: not a TIFF file (header b'' not valid)

#38 lunchben closed 4 years ago
1
Typo error.

#37 mo-vic opened 4 years ago
0
About the figure of training the model

#36 SWHL opened 5 years ago
0
谢谢作者的无私分享。我想自己制作数据集用此网络试试效果，请问我该怎么标注图片？

#35 lunchben closed 4 years ago
6
我只有一张6G的显卡，bachsize设置为4，训练时测试的正确率一直徘徊在0.6到0.7上不去，这是为什么啊？

#34 rechealzrq opened 5 years ago
1
proposal对齐问题

#33 he010103 opened 5 years ago
7
ImportError: cannot import name 'Car2000'

#32 Totoro-wen opened 5 years ago
0
测试问题！！！

#31 gengyanlei opened 5 years ago
21
图片读取时的问题

#30 lisili opened 5 years ago
0
你的代码有个小问题！！！

#29 gengyanlei closed 3 years ago
2
consult/请教

#28 gengyanlei opened 5 years ago
0
怎么感觉代码中实现rank loss和论文中描述的不一样

#27 yongqis closed 5 years ago
1
为什么每块GPU里面batchsize数量越少，效果越好，8 img per gpu只有86.2，但是4 img per gpu能到87.1

#26 tiancity-bytedance opened 5 years ago
1
请问论文中的图5是怎么得到的矩形框？

#25 tectal opened 5 years ago
3
feature extraction

#24 Shahnawazgrewal opened 5 years ago
0
Returning the image crops?

#23 muellerzr opened 5 years ago
0
_, term_width = os.popen('stty size', 'r').read().split()

#22 tectal opened 5 years ago
4
List Loss error, IndexError: index 4 is out of bounds for dimension 0 with size 4

#21 muellerzr closed 5 years ago
3
cub200-2011，config文件中的参数设置没变batch=16，双卡训练，准确率最高在0.869。

#20 weihua93 closed 5 years ago
14
无法下载你们训练的模型

#19 xiaoguiziwandao opened 5 years ago
0
The way to compute rank loss

#18 yaojunr opened 5 years ago
3
在standford cars 数据集上，测试正确率88.6%左右就上不去了，有达到论文中的93.9%的吗，batch size为8

#17 dongzhi0312 opened 5 years ago
20
Is it possible to use ResNext?

#16 zzzucf closed 5 years ago
1
What is PROPOSAL_NUM = 6

#15 gabrielzhuyun opened 5 years ago
1
dropout in resnet.py

#14 uniquezhengjie opened 5 years ago
2
矩阵大小不一致

#13 zhaozhiyong19890102 closed 5 years ago
0
need 4 optimizer ？

#12 shiyongde opened 5 years ago
1
stty: 'standard input': Inappropriate ioctl for device

#11 ray-lee-94 closed 5 years ago
6
Anchor boxes visualization

#10 ygean opened 5 years ago
9
大家测试过吗，结果和论文里的匹配吗

#9 zbxzc35 opened 5 years ago
17
test

#8 sankin1770 closed 5 years ago
0