kongbia

refine forward

2

``` if test: p0 = torch.nn.functional.pad(f[0], [16, 16, 16, 16])[:, :, 4*pos[0]:4*pos[0]+61, 4*pos[1]:4*pos[1]+61] p1 = torch.nn.functional.pad(f[1], [8, 8, 8, 8])[:, :, 2 * pos[0]:2 * pos[0] + 31, 2 *...

个人博客中的域名变化了一次，导致旧的评论无法显示。我试过将issue的label改成一样的，似乎没有用，请问有解决方法吗？ 旧域名的评论： https://github.com/kongbia/kongbia.github.io/issues/11 新域名的评论： https://github.com/kongbia/kongbia.github.io/issues/42

运行错误

2

在测试demo_rh_pyramid.py中出现如下错误 Traceback (most recent call last): File "demo_rh_pyramid.py", line 393, in args.h_overlap, args.w_overlap, args.save_res) File "demo_rh_pyramid.py", line 69, in inference mask_batch=None) File "../libs/networks/build_whole_network.py", line 447, in build_whole_detection_network img_shape=img_shape) File "../libs/networks/build_whole_network.py",...

test multiple algorithm

2

How to test multiple algorithms simultaneously using multiple GPUs?

TypeError about Analysis vot2020

5

when I analysis the result following the official guidelines ``` vot analysis --workspace --output json ``` it meets a TypeError "TypeError: Object of type 'Grid' is not JSON serializable" Can...

about lmdb

1

Hi，How to make dataset in lmdb format? Would it accelerate training?

raw results

2

Could you please share the raw results of VOT2018 and LaSOT

VOT2018结果

3

我用提供的raw results结果在vot2018测试，offline EAO是0.5020，online EAO只有0.4389

Could you provide the original result files

2

Could you please provide the original result files for the convenience of ploting curves like OTB,VOT,GOT10k?

got10k训练结果

11

您好，我在4张titanv上训练got10k， batch32，初始学习率扩大两倍，epoch 总共500，在epoch 250下降学习率，最后结果只有64.6。请问如何该配置？