zhuzhu comments

Results 12 comments of


                                            zhuzhu

Dolly Zoom effect

Have you implemented the dolly zoom effect? how did you do it?

What's the correspondence of joints in Posetrack format to those in COCO format?

> > > > > I am also interested in evaluating the model on posetrack data. > > > > > > > > > @nikhilaravi > > > Actually...

Thank you for your sharing！ Is there a training code? I hope I can train my own dataset to see the effect

Have you reproduce train code now?

Could not find a version that satisfies the requirement splatting

you can find the "splatting" module here: https://github.com/hperrot/splatting

> line [here](https://github.com/Sampson-Lee/myDRGAN/blob/f55f61de884b90543117a0768a2e04c9182cec4d/run/train_single_DRGAN.py#L87) I think it should be: > > ```python > batch_pose_code = torch.FloatTensor(batch_pose_code.tolist()) > ``` I find it should be so. Is it a bug?

size issue on GAE process

[see here](https://github.com/uvipen/Super-mario-bros-PPO-pytorch/issues/9)

about train.py

I think it should be as you

TNN convert & Android inference Error: can't infer resource shape from binary param in benchmark mode, random generator may not be exactly same with the real resource. Segmentation fault.

我也出现了同样的问题，怎么解决

TNN convert & Android inference Error: can't infer resource shape from binary param in benchmark mode, random generator may not be exactly same with the real resource. Segmentation fault.

> 我也出现了同样的问题，怎么解决我发现是pytorch模型在转换成ONNX时将部分Constant算子折叠了，导致在ONNX转成tnnmodel时常量算子丢失，缺失了部分算子，无法推理

ModuleNotFoundError: No module named 'realesrgan.version'

> Follow steps from https://github.com/xinntao/Real-ESRGAN#installation 并不想直接安装这个库，想在服务器直接跑