profile
viewpoint
tomguluson92 Beijing, China small tommy

tomguluson92/color_recognition 1

color recognition by samuel

tomguluson92/3DDFA 0

The PyTorch improved version of TPAMI 2017 paper: Face Alignment in Full Pose Range: A 3D Total Solution.

tomguluson92/attention 0

some attention implements

tomguluson92/auto-email 0

a easier and settable email-sender with multiple lines

tomguluson92/benchmarking-gnns 0

Repository for benchmarking graph neural networks

tomguluson92/Bitcoin-Trader-RL 0

A profitable cryptocurrency trading environment using deep reinforcement learning and OpenAI's gym

tomguluson92/blender-cli-rendering 0

Python scripts for rendering images using Blender 2.80+ from command-line interface

tomguluson92/bootstrap 0

The most popular HTML, CSS, and JavaScript framework for developing responsive, mobile first projects on the web.

startedAntixK/PyTorch-VAE

started time in a day

startedpodgorskiy/DareBlopy

started time in a day

startedcharleschiu2012/FineTune

started time in 2 days

fork tomguluson92/curated-list-of-awesome-3D-Morphable-Model-software-and-data

The idea of this list is to collect shared data and algorithms around 3D Morphable Models. You are invited to contribute to this list by adding a pull request. The original list arised from the Dagstuhl seminar on 3D Morphable Models https://www.dagstuhl.de/19102 in March 2019.

fork in 5 days

issue commenttomguluson92/StyleGAN_PyTorch

the LeakyReLU layer after dense1

Thanks for pointing that problem, do you found that useful if removing this?

Walstruzz

comment created time in 6 days

startedalibaba/MNN

started time in 8 days

startedryfeus/tensorflow2pytorch

started time in 8 days

startedmitsuba-renderer/mitsuba2

started time in 9 days

issue commentmicrosoft/Deep3DFaceReconstruction

[Reconstruction Result] wrong direction

Hi @YuDeng, does R(x) in computing texture variance loss is the skinmask in BFM_model_front.mat? Since I found I always get a mean face, I think maybe Ltex is essential for a pleased result?

image

tomguluson92

comment created time in 9 days

startedxingmimfl/Deep3DFacereconstruction_pytorch3d

started time in 10 days

issue commentmicrosoft/Deep3DFaceReconstruction

[Reconstruction Result] wrong direction

Thank you so much, I will do that and examine it.

tomguluson92

comment created time in 10 days

issue commentmicrosoft/Deep3DFaceReconstruction

[Reconstruction Result] wrong direction

Dear Dr. Deng, I just wonder if the result is reasonable (at the first stage of running, I only use 100 pics as dataset for training). Besides, could you mind give me the specific meaning of R(x) in Equation 6? Would that affects big in the final output? The loss image

The result image

tomguluson92

comment created time in 11 days

PR opened google/tf_mesh_renderer

Fix the bug of edge_over_w

Hi there, According to the derivative rules, I found edge1_over_w should equal to x0 * db1_dx + y0 * db1_dy + w0 * db1_dw. As edge function defined in <Olano & Greer, 1997> said: Ei(X, Y, W) = (X-Xi)dXi + (Y-Yi)dYi + (W-Wi)dWi, triangle is ordered as anti-clockwise like below

image

On top of these evidence, I believe this implementation is wrong, and I have test it with your test_case file, it's passes backward gradient verification.

+1 -1

0 comment

1 changed file

pr created time in 11 days

push eventtomguluson92/tf_mesh_renderer

tomguluson92

commit sha 6710a34ee38fe12bc100090dc1ec68481e83eb46

Fix the bug of edge_over_w Hi there, According to the derivative rules, I found `edge1_over_w` should equal to `x0 * db1_dx + y0 * db1_dy + w0 * db1_dw`. As edge function defined in <Olano & Greer, 1997> said: `Ei(X, Y, W) = (X-Xi)dXi + (Y-Yi)dYi + (W-Wi)dWi`, triangle is ordered as anti-clockwise like below ``` v0 / \ / \ / \ v2 ------ v1 ``` On top of these evidence, I believe this implementation is wrong, and I have test it with your test_case file, it's passes backward gradient verification.

view details

push time in 11 days

fork tomguluson92/tf_mesh_renderer

A differentiable, 3D mesh renderer using TensorFlow.

fork in 11 days

issue commentmicrosoft/Deep3DFaceReconstruction

[Reconstruction Result] wrong direction

Thank you very much!

tomguluson92

comment created time in 12 days

issue commentmicrosoft/Deep3DFaceReconstruction

[Reconstruction Result] wrong direction

Thanks a lot! I found it's may caused by I did not use a pretrianed resnet50 at first.

Does this kind of results seems reasonable? By the way, which normalization technique do you use?

normalize the data to 0 to 1 or -1 to +1 ?

image

tomguluson92

comment created time in 12 days

issue openedmicrosoft/Deep3DFaceReconstruction

[Reconstruction Result] wrong direction

Dear Deng, I have implement a pytorch version of your repo (including tf_mesh_renderer). But during training process (I only use photonmeteric loss, perceptual loss and coefficient regularization loss), I got the reconstructed faces with wrong direction. Any ideas on solving this? Really appreciated for your response! image

created time in 13 days

startedanandpawara/Real_Time_Image_Animation

started time in 17 days

startedzllrunning/face-parsing.PyTorch

started time in 17 days

startednileshkulkarni/acsm

started time in 18 days

startedweihaox/awesome-neural-rendering

started time in 19 days

issue commenttomguluson92/unsup3D_pytorch3d

Excellent work!

Nope, But I think this is due to the different rendering interface. I think one could get a reasonable result with dive more into the code of unsup3D. I think my version is just not enough,

raywzy

comment created time in 19 days

startedczl0325/jetbrains-agent-latest

started time in 20 days

fork tomguluson92/pytorch-tvmisc

Totally Versatile Miscellanea for Pytorch

fork in 21 days

startedt-vi/pytorch-tvmisc

started time in 21 days

startedzhipeng-fan/3D-face

started time in a month

PR opened pytorch/tutorials

fix deprecated usage `ctx.saved_variables`

update ctx.saved_variables of torch.autograd.Function to ctx.saved_tensors.

+1 -1

0 comment

1 changed file

pr created time in a month

push eventtomguluson92/tutorials

tomguluson92

commit sha 1acfec518533b6d73c2d82d3664cb57e7736436a

fix deprecated usage `ctx.saved_variables` update **ctx.saved_variables** of `torch.autograd.Function` to **ctx.saved_tensors**

view details

push time in a month

startedsuphoff/pytorch_parallel_extension_cpp

started time in a month

startedgoogle/tf_mesh_renderer

started time in a month

startedgradio-app/gradio

started time in a month

startedaliyun/oss-browser

started time in a month

startedgenforce/idinvert

started time in a month

startedvsitzmann/siren

started time in a month

issue commentTencent/TNN

转换tensorflow的pb模型出错

可以的。你发上来,我来看下

哈喽,请问您试验结果如何?

tomguluson92

comment created time in a month

startedzxdefying/pytorch_tricks

started time in a month

startedwtjiang98/PSGAN

started time in 2 months

startedsainnhe/dotfiles

started time in 2 months

push eventtomguluson92/StyleGAN_PyTorch

tomguluson92

commit sha 4fd0711f560b9b080fca3df2448822835226ba02

Update README.md

view details

push time in 2 months

push eventtomguluson92/StyleGAN_PyTorch

tomguluson92

commit sha b38ee3e16ea8688d97b639f4360fc1e94d199d30

Update README.md

view details

push time in 2 months

push eventtomguluson92/StyleGAN_PyTorch

tomguluson92

commit sha 6d75a91d0de8516770fd2f6006b708ff4c0829f4

Update README.md

view details

push time in 2 months

push eventtomguluson92/StyleGAN_PyTorch

tomguluson92

commit sha bfc342558c0b00a3ad3ea32e553cc6852dab3e18

Create FUNDING.yml

view details

push time in 2 months

push eventtomguluson92/unsup3D_pytorch3d

tomguluson92

commit sha 29690451cd6edf435406c8eac0e240a208b89553

Update README.md

view details

push time in 2 months

push eventtomguluson92/unsup3D_pytorch3d

tomguluson92

commit sha 7831e71af51e1806236f5882a29d030029dbcbfd

update readme

view details

push time in 2 months

issue commentTencent/TNN

转换tensorflow的pb模型出错

模型地址: https://c-t.work/s/ce622bfa78cb4f (提取码: hbgdkm)

特殊层: PixelShuffle(本质上是tf.depth_to_width, 即只负责特征图维度的转换).

转换命令:

python3 converter.py tf2tnn \
    -tp l256.pb \
    -in "input0[1,3,256,256]" \
    -on model_2/conv2d_15/Sigmoid, model_2/conv2d_10/Sigmoid\
    -v v2.0 \
    -optimize \
    -o ~/tf-model/ \
    -align \
tomguluson92

comment created time in 2 months

issue commentTencent/TNN

转换tensorflow的pb模型出错

和您确认下? 现在可以转换了吗?

非常感谢, 还是不行呀。我可以把模型和转换命令发您吗?

tomguluson92

comment created time in 2 months

startedCalciferZh/SMPL

started time in 2 months

push eventtomguluson92/unsup3D_pytorch3d

tomguluson92

commit sha 1adbcfe2f2934e4981ea04f8855ae54004ad041d

Update README.md

view details

push time in 2 months

push eventtomguluson92/unsup3D_pytorch3d

tomguluson92

commit sha 27b4d6757dd631aa35631c147c81bda3293b57ad

update readme & wrong camera settings

view details

tomguluson92

commit sha 3a1983d3341649e531d658230e2661ddcccfdf71

Merge branch 'master' of https://github.com/tomguluson92/unsup3D_pytorch3d

view details

push time in 2 months

push eventtomguluson92/unsup3D_pytorch3d

tomguluson92

commit sha 3376c2b726da4e4fb9d82e69febff682df72a956

Update renderer.py

view details

push time in 2 months

issue commentelliottwu/unsup3d

replace neural renderer with softRas

Do you have wechat or something? I have some question about the detail and want to dicuss with you. my wechat is:

goodaihong

tomguluson92

comment created time in 2 months

push eventtomguluson92/unsup3D_pytorch3d

tomguluson92

commit sha 3118a7cce4d72be7d8578e3b58caaceaf86867f1

Update utils.py

view details

push time in 2 months

issue commentelliottwu/unsup3d

replace neural renderer with softRas

I see! Have you tried to train with SoftRas on the full training set? Does it still not work well?

I will give it a try and tell you my result.

tomguluson92

comment created time in 2 months

issue commentelliottwu/unsup3d

replace neural renderer with softRas

Hi! First of all, thank you for putting efforts in this! I'm keen to learn about this as well.

I have not tried SoftRas yet, but I would not expect it too be far worse. It can be slightly worrisome, since SoftRas softens the forward rendering to smooth the gradients, which can also lead to inconsistency and provide noisy signals for the shape. That said, I don't expect it to fail completely.

Neural Mesh Renderer (NMR) on the other hand performs exact forward rendering. However, we found gradients through the texture are noisy and unstable for training. Therefore we only use it to render the depth and use grid_sample to warp the texture, as explained in the supplementary material.

However, when I look at the results you show in your repo, it seems you were not even able to reproduce the results with NMR? Did you change anything? If not, have you trained the model sufficiently long (eg, 12+h)?

I may not be able to look into SoftRas recently, but I think investigating this issue can be rather helpful for many users. So keep us posted with any updates!

Thanks! The poor results may be due to the choosen of a small dataset which consists of only 40 pictures to trained on (celeba is big for me to test the effects).

tomguluson92

comment created time in 2 months

startedtomguluson92/unsup3D_pytorch3d

started time in 2 months

push eventtomguluson92/unsup3D_pytorch3d

tomguluson92

commit sha 0930c7e9cecacfeec1926da9e17e77a4d60d5c03

unsup3d pytorch3d(softRas)

view details

push time in 2 months

issue openedelliottwu/unsup3d

replace neural renderer with softRas

Hi, wu! congratulations to you for unsup3d has been elected as the CVPR2020 best paper. Inspired from your repo, I just replace the neural renderer with pytorch3d point cloud renderer. My repo is: https://github.com/tomguluson92/unsup3D_pytorch3d But I found it is inferior to your repo, but as far as I know that SoftRas inside pytorch3d is a more powerful differentiable renderer. Do you have time in using pytorch3d and find out the difference between them? Thanks a lot!

created time in 2 months

create barnchtomguluson92/unsup3D_pytorch3d

branch : master

created branch time in 2 months

created repositorytomguluson92/unsup3D_pytorch3d

created time in 2 months

startedtuna/tunasync-scripts

started time in 2 months

startedelliottwu/unsup3d

started time in 2 months

startederguotou520/bye

started time in 2 months

startedFuxiCV/3D-Face-GCNs

started time in 2 months

startedsaic-violet/textured_avatars

started time in 2 months

startedCivilNet/Gemfield

started time in 2 months

issue commentLansburyCH/relightable-nr

What is the meaning of `relightable` in Neural Rendering?

Thanks for your interest in our work. For free viewpoint video, it synthesizes novel views of a dynamic scene without changing its lighting condition. As such, the synthesized appearance is limited to be under the lighting condition of the input video sequences. On the contrary, our work can synthesize the appearance under not only novel views but also novel lighting (but currently only for static objects). This second task is often referred to as relighting, and that is where relightable comes from. Naturally, we can also carry out a combination of the two tasks, which is usually referred to as free-viewpoint relighting in previous literature. The reason for associating relightable with Neural Rendering is that current neural rendering (rerendering) methods mostly focus on the change of viewpoints, but do not address much about the change of lighting, which is also an important part.

Thanks bro, much appreciated!

tomguluson92

comment created time in 2 months

startedmtyylx/MLND

started time in 2 months

startedzhan-xu/RigNet

started time in 2 months

issue openedLansburyCH/relightable-nr

What is the meaning of `relightable` in Neural Rendering?

Hi,

Thanks a lot for your great work! 
I am curious about the meaning of `relightable` and its relationship with `free viewpoint video`?

created time in 2 months

startedLansburyCH/relightable-nr

started time in 2 months

startedlightaime/sgas

started time in 2 months

issue openedTencent/TNN

转换tensorflow的pb模型出错

Hi, 非常感谢腾讯大佬们的TNN!

我这里遇到一点小问题来到这里请教, 报错信息如下:

...
weight = const_fold_opt__100
weight = Const__82
python3: xxxx/onnx_utility.cc:544: const float* get_tensor_proto_data(const onnx::TensorProto&): Assertion `0` Aborted (core dumped).

根据我的分析,因为assert(0)报错,而在onnx_utility.cc的544行的printf语句并没把对应的数据类型打印到控制台, 我改用std::cout也没有打印到控制台.

请问如何去查看printf的结果呢?以便检查是哪方面的问题?

created time in 2 months

PR opened Tencent/TNN

Fix tf2tnn output dir problems
+2 -2

0 comment

1 changed file

pr created time in 2 months

push eventtomguluson92/TNN

tomguluson92

commit sha 4f2efdbedf47266d1ec63057d01cfac3da8ce420

Fix tf2tnn output dir problems

view details

push time in 2 months

fork tomguluson92/TNN

TNN:由腾讯优图实验室打造,移动端高性能、轻量级推理框架,同时拥有跨平台、高性能、模型压缩、代码裁剪等众多突出优势。TNN框架在原有Rapidnet、ncnn框架的基础上进一步加强了移动端设备的支持以及性能优化,同时也借鉴了业界主流开源框架高性能和良好拓展性的优点。目前TNN已经在手Q、微视、P图等应用中落地,欢迎大家参与协同共建,促进TNN推理框架进一步完善。

fork in 2 months

startedmicrosoft/DisentangledFaceGAN

started time in 2 months

startedNVIDIA/vid2vid

started time in 2 months

startedlturing/tacotronv2_wavernn_chinese

started time in 2 months

startedTencent/TNN

started time in 2 months

startedlayumi/person-reid-3d

started time in 2 months

startedBreakthrough/PySceneDetect

started time in 2 months

startedpvjosue/pytorch_convNd

started time in 2 months

startedchangkun/modern-cpp-tutorial

started time in 2 months

fork tomguluson92/phkit

phoneme toolkit. 好用的音素处理工具箱,包含中文音素、英文音素、文本转拼音、文本正则化等模块。

fork in 2 months

startedKuangDD/phkit

started time in 2 months

startedserengil/tensorflow-101

started time in 2 months

startedprotossw512/AdaptiveWingLoss

started time in 2 months

startedsniklaus/3d-ken-burns

started time in 2 months

startedgoogle/mediapipe

started time in 2 months

fork tomguluson92/tacotron2-1

Forked from NVIDIA/tacotron2 and merged with Rayhane-mamah/Tacotron-2

fork in 2 months

startedbegeekmyfriend/tacotron2

started time in 2 months

issue openedRudrabha/Lip2Wav

Lip2Wav dataset not available

Hi,

I am really happy to see your pioneering work! And I am wondering whether can I download the Lip2Wav dataset? (the project website do not offer the URL).

Really appreciate it, thank you~

created time in 3 months

startedJasonWei512/Tacotron-2-Chinese

started time in 3 months

startedopenai/gpt-3

started time in 3 months

startedEricGzq/Hybrid-Fake-Face-Dataset

started time in 3 months

startedpclucas14/pytorch-glow

started time in 3 months

startedpseeth/torch-stft

started time in 3 months

more