pengboxiangshang / siamrpn_plus_plus_pytorch Goto Github PK

View Code? Open in Web Editor NEW

475.0 18.0 90.0 766 KB

SiamRPN, SiamRPN++, unofficial implementation of "SiamRPN++" (CVPR2019), multi-GPUs, LMDB.

Python 99.42% Shell 0.58%

siamrpn_plus_plus_pytorch's People

Contributors

Stargazers

Watchers

Forkers

johndpope bytedancers kelvinson hajungong007 dreadlord1984 cvrosefun shiyongde leo-xxx bruinxiong yangkang779 cclauss 983632847 chiukin jacke121 saintlogos1234 13331112522 lkn123 peterzs xuhuaze707313 lc4321 wjgaas zhaomonica llf10811020205 zhushansheng songchengwen 10183308 gaocode xiaoyubing jstzwjr meitianjinbu sengbupt aobai hyzcn silencesss column6942 bigbugx cxsunshine collector-m nculsz qtjiebin fengsijia peterzhousz haoliyoupai09 ricardomlyk batermj lfs119 lullabyafa perfectzh ason93 chocolateqin hhgxx123 ll752003446 fmsunyh rodrigogantier wulingtian jadentan doublepoints donghaw caimingxie ghostlyfeng koala7580 zj717754140 tonylibing tqcai belye guo-kaijun medical-utralsonic-object-detection alei089 labimage youtang1993 turgunyusuf ingeniousfrog xiaohu-d hanwsf liuzhongshen dongzhou-1996 guang-yao llll-hui xieyunjiao shartoo alvin-gys freegliboracle veryquant nuaasxr ssbagalkar shellylingling zuzi2015 t-xi sabadijou veritasxu

siamrpn_plus_plus_pytorch's Issues

pretrained model

Can you provide pretrained model?

Thanks!

Do you know the loss of siamrpn++ in detail?

请教

您好，您这边要求的Ubuntu 14.04、Python 2.7、PyTorch 0.4.0
我想请教下，我使用Ubuntu 17.10、Python3、PyTorch 0.4的话是否可以？

about auc

Great job!
I am wondering what is the auc on otb and vot?

the question about random_shift_factor

Which code handles the sampling problem of random_shift_factor=32？

When will the test code be published?

When will the test code be published, only training can't test?

can you please tell your result in

can you tell me your result in otb and vot2018

How much lmdb format training can improve the training speed?

Excuse me, I want to know how much the training speed improving when using the datasets with the lmdb format.

Transfer Learning

Hi.
Can you please share some transfer learning strategy to start with
without training from scratch?
thank you

worse result for test auc in otb100

thx for this great job at first！！！
i trained this code ONLY in ILSVRC2015_finetune,not like mult-trained(eg: trained in COCO,YoutubebBB，ILSVRC2015) in paper,and i got the auc:0.50 for epoch 20 in OTB100 testset.It's much worse than the results in the paper（0.69）.
so Reasons for this phenomenon should be Too little training data？（only in ILSVRC2015）

where is the raw results?

Hi, could you please give me the raw results? Thanks!

About the IOU

Thank you for your work,but I have a problem about the IOU , as followed:

cy_o = (im_h - 1.) / 2
cx_o = (im_w - 1.) / 2
cy = cy_o + np.random.randint(- self.max_translate, self.max_translate + 1)
cx = cx_o + np.random.randint(- self.max_translate, self.max_translate + 1)
gt_cx = cx_o - cx
 gt_cy = cy_o - cy

I don't understand these codes,can someone explain why the ground truth of the object is written as above? Thank you .