Darius-H

你还在使用被淘汰的www.weibo.cn网址，该网页已经弃用了。爬取信息很麻烦，也不全

cleverhans.torch.utils.clip_eta will cause GPU memory leaking for its in-place operation

1

cleverhans.torch.utils.clip_eta will cause GPU memory leaking for its in-place operation when param norm==2 ``` def clip_eta(eta, norm, eps): """ PyTorch implementation of the clip_eta in utils_tf. :param eta: Tensor :param...

Image is resized when using tool 'sketch'

2

### Describe the bug I have an image with shape (height=80,width=848), and I want to use the 'sketch' tool to generate mask for it. However, whether or not I use...

Convert weakperspective camera to perspective camera

1

I'm trying to run https://github.com/open-mmlab/mmhuman3d/blob/main/demo/pymafx_estimate_smplx.py to get smplx params and perspective camera matrices. I tried to extract the weakperspective camera matrices R,T,K from the code here https://github.com/open-mmlab/mmhuman3d/blob/9431addec32f7fbeffa1786927a854c0ab79d9ea/mmhuman3d/core/visualization/visualize_smpl.py#L998C1-L1007C27 and covert them...

I find that when using pytorch3d, convert_screen_to_ndc function does not work properly, it generates a wrong principal point. The convert_screen_to_ndc is here: https://github.com/open-mmlab/mmhuman3d/blob/9431addec32f7fbeffa1786927a854c0ab79d9ea/mmhuman3d/core/conventions/cameras/convert_convention.py#L481C1-L481C1 I find that in pytorch3d’s code ,...

Is adaLN applicable to text condition?

3

In DiTBlock: self.adaLN_modulation = nn.Sequential( nn.SiLU(), nn.Linear(hidden_size, 6 * hidden_size, bias=True) ) def forward(self, x, c): shift_msa, scale_msa, gate_msa, shift_mlp, scale_mlp, gate_mlp = self.adaLN_modulation(c).chunk(6, dim=1) x = x + gate_msa.unsqueeze(1)...

What's the meaning of bones' numbers in deformer?

1

https://github.com/xuchen-ethz/snarf/blob/ae0c893cc049f0f8270eaa401e138dff5d4637b9/lib/model/deformer.py#L29 What's the meaning of these bones' numbers? Any reference available?

Where to download the melception.pt to evaluate?

2

In Codebook/readme.md, it said we can run python Codebook/evaluate.py to test the metrics. However it needs /apdcephfs/share_1316500/donchaoyang/code3/Codebook/evaluation/logs/melception.pt to extract features. I didn't find the checkpoint in this project, neither the...

About the experiment data

2

Dear author: Can you provide the data download link? And did you make the train json file by yourself? "train_data": "data/charades_sta/train.json"

纯数有问题

1

print(Text(raw_text='纯数：2983.07克或12345.60米。').normalize()) 结果是 纯数：二九八三.零七克或一二三四五.六十米。 点没有正确识别 需要把https://github.com/atomicoo/chn_text_norm/blob/8210575e9a4fddb409e5a76b922bf6c30f5833c5/text.py#L114-L120 规范化数字编号挪到规范化纯数后面