profile
viewpoint
Yifan Wang yifita ETH Zurich Zurich https://yifita.github.io PhD student @ ETH Zurich

fperazzi/proSR 879

Repository containing an independent implementation of the paper: "A Fully Progressive Approach to Single-Image Super-Resolution"

yifita/DSS 143

Differentiable Surface Splatting

yifita/3PU 118

Patch-base progressive 3D Point Set Upsampling

yifita/deep_cage 70

code for "Neural Cages for Detail-Preserving 3D Deformations"

yifita/3PU_pytorch 46

pytorch implementation of >>Patch-base progressive 3D Point Set Upsampling<<

yifita/action.sr_cnn 35

code for Two-Stream SR-CNNs for Action Recognition in Videos 2016

yifita/pytorch_points 16

a pytorch extension for learning point clouds

swook/SfM 10

Structure from Motion pipeline implementation using RGB-D images

swook/DynamicPeople 4

Our solutions for the course "Dynamic Programming and Optimal Control", autumn semester 2014

issue closedyifita/deep_cage

multifold gen

Hi,

Thanks for the great work! I was wondering the meaning of "MultiFoldPointGen", since you only use the last fold. Could you please explain that?

Thanks,

closed time in 15 days

Colin97

issue commentyifita/deep_cage

multifold gen

Hi Colin,

Thanks for paying such close attention. To me, it meant to refer to the fact that the network has multiple "folds".

Colin97

comment created time in 15 days

issue closedyifita/deep_cage

Why use weights = weights.detach() on line 137 of the file common.py?

Hi, thanks for your excellent results on this paper and the published code. I try to do some experiment with your code. I successfully train the network, but I have a little doubt.

On line 137 of the file common.py , you used: weights = weights.detach(). And in file losses.py, you used: *self.loss["WREG"]+=self.mvc_reg_loss(all_outputs["weight"])self.opt.mvc_weight While all_outputs[“weight”].requires_grad == False here, I am confused about whether loss WREG affects the final result. From my experiment , I find that no matter how big the value of mvc_weight is, the result has not changed.

Looking forward to your reply, thank you.

closed time in 15 days

xiruiYan

push eventyifita/deep_cage

Yifan Wang

commit sha da4c07dcc5c10d44c1b040fbbd0d897ea5ab1f98

Update common.py remove weights.detach()

view details

push time in 15 days

issue commentyifita/deep_cage

Why use weights = weights.detach() on line 137 of the file common.py?

Hi Xirui,

Well spotted, thank you! After checking the dev repo, I found this line in common.py was indeed added later, out of context. I'm going to comment it out.

Thank you very much for point it out!!

xiruiYan

comment created time in 15 days

push eventyifita/deep_cage

ywang

commit sha 2e9e09e33456451abafd05a0806fe7fb201de7a0

put processed data igl

view details

push time in 16 days

push eventyifita/pytorch3d

Nikhila Ravi

commit sha 5852b74d122f95fc7eeb041cd11593f647b9fc7a

Softmax blending small fix Summary: Small fix to the softmax blending function. To avoid overflow in the exponential for the softmax, the exponent is shifted by the maximum value. In the final calculation of the color there is a weighted sum between the pixel color and the background color - in order for the sum to be correct, the background color also needs to be handled in the same way witt the shifted exponent. Reviewed By: gkioxari Differential Revision: D23148301 fbshipit-source-id: 86066586ee7d3ce7bd4a2076b12ce191fbd151a7

view details

Georgia Gkioxari

commit sha 7f2f95f2255431267b9aabf90bb2bba7bb2b0880

detach for meshes, pointclouds, textures Summary: Add `detach` for Meshes, Pointclouds, Textures Reviewed By: nikhilaravi Differential Revision: D23070418 fbshipit-source-id: 68671124ce114c4495d7ef3c944c9aac3d0db2d8

view details

root@sandcastle5743.frc3.facebook.com

commit sha f03aa5803bcdba3b9447a8c2ffdc77e693d51c87

suppress errors in `vision/fair/pytorch3d` Differential Revision: D23180198 fbshipit-source-id: cad1fa7ba9935f3ca20410a2575e173999c04be1

view details

Anton Troynikov

commit sha 370f1c380c3f025103f88fce1eaf0e0d990eb7e4

Fix 404 link in Renderer Doc (#330) Summary: Because of the way Sphinx was parsing this link in Markdown, the link wasn't working properly. This should fix it. Pull Request resolved: https://github.com/facebookresearch/pytorch3d/pull/330 Test Plan: Tested via local Sphinx. Reviewed By: nikhilaravi Differential Revision: D23244163 Pulled By: atroyn fbshipit-source-id: 019712a841d76391a5210dcd98c77a822947204a

view details

mlygao@devgpu002.atn3.facebook.com

commit sha 9242e7e65d3c1c841333ece7b7469cd677100e34

ShapeNetCore & R2N2 tutorial Summary: Tutorial for ShapeNetCore & R2N2. Reviewed By: gkioxari Differential Revision: D22916882 fbshipit-source-id: 752742be87f44919164ec7eafcc9c09c17a0f8a3

view details

Georgia Gkioxari

commit sha 57a22e7306db24140cb2133aa12a613cbf971c4c

camera refactoring Summary: Refactor cameras * CamerasBase was enhanced with `transform_points_screen` that transforms projected points from NDC to screen space * OpenGLPerspective, OpenGLOrthographic -> FoVPerspective, FoVOrthographic * SfMPerspective, SfMOrthographic -> Perspective, Orthographic * PerspectiveCamera can optionally be constructred with screen space parameters * Note on Cameras and coordinate systems was added Reviewed By: nikhilaravi Differential Revision: D23168525 fbshipit-source-id: dd138e2b2cc7e0e0d9f34c45b8251c01266a2063

view details

Steve Branson

commit sha 9aaba0483c08c9a40c26db0858f8c0688f33e850

Temporary fix for mesh rasterization bug for traingles partially behind the camera Summary: A triangle is culled if any vertex in a triangle is behind the camera. This fixes incorrect rendering of triangles that are partially behind the camera, where screen coordinate calculations are strange. It doesn't work for triangles that are partially behind the camera but still intersect with the view frustum. Reviewed By: nikhilaravi Differential Revision: D22856181 fbshipit-source-id: a9cbaa1327d89601b83d0dfd3e4a04f934a4a213

view details

Jeremy Reizenstein

commit sha 9a50cf800e511031602cb4b9523cef7b448b16af

Fix batching bug from TexturesUV packed ambiguity, other textures tidyup Summary: faces_uvs_packed and verts_uvs_packed were only used in one place and the definition of the former was ambiguous. This meant that the wrong coordinates could be used for meshes other than the first in the batch. I have therefore removed both functions and build their common result inline. Added a test that a simple batch of two meshes is rendered consistently with the rendering of each alone. This test would have failed before. I hope this fixes https://github.com/facebookresearch/pytorch3d/issues/283. Some other small improvements to the textures code. Reviewed By: nikhilaravi Differential Revision: D23161936 fbshipit-source-id: f99b560a46f6b30262e07028b049812bc04350a7

view details

Nikhila Ravi

commit sha d330765847435b982eec0da7b653044e86905fcb

Website and docs updates Summary: - Added sbranson's fit mesh tutorial to the website - Updated rendering docs with info about texturing and new shader types. TODO: - add pointcloud rendering tutorial to the website as well (https://github.com/facebookresearch/pytorch3d/blob/master/docs/tutorials/render_colored_points.ipynb) - docs for camera - update some tutorials which depended on the Textures from structures. Reviewed By: gkioxari Differential Revision: D23143977 fbshipit-source-id: 6843c9bf3ce11115c459c64da5b0ad778dc92177

view details

Nikhila Ravi

commit sha 90f6a005b09af9570fc50799a301303e49b67dfe

Tutorials textures updates and fix bug in extending meshes with uv textures Summary: Found a bug in extending textures with vertex uv coordinates. This was due to the padded -> list conversion of vertex uv coordinates i.e. The number of vertices in the mesh and in verts_uvs can differ e.g. if a vertex is shared between 3 faces, it can have up to 3 different uv coordinates. Therefore we cannot convert directly from padded to list using _num_verts_per_mesh Reviewed By: bottler Differential Revision: D23233595 fbshipit-source-id: 0c66d15baae697ead0bdc384f74c27d4c6539fc9

view details

Nikhila Ravi

commit sha 778383eef77a23686f3d0e68834b29d6d73f8501

Texture loading and rendering in ShapeNetCore and R2N2 data loaders Summary: - Add support for loading textures from ShapeNet Obj files as a texture atlas. - Support textured rendering of shapenet models Reviewed By: gkioxari Differential Revision: D23141143 fbshipit-source-id: 26eb81758d4cdbd6d820b072b58f5c6c08cb90bc

view details

Eduardo Henrique Arnold

commit sha d0cec028c7ddb59b8e4d0ee054d0ea8b1a493d46

Fix look_at_view_transform when object location is not (0,0,0) (#230) Summary: The look_at_view_transform did not give the correct results when the object location `at` was not (0,0,0). The problem was on computing the cameras' location in world's coordinate `C`. It only took into account the camera position from spherical angles, but ignored the object location in the world's coordinate system. I simply modified the C tensor to take into account the object's location which is not necessarily in the origin. I ran unit tests and all but 4 failed with the same error message: `RuntimeError: CUDA error: invalid device ordinal`. However the same happens before this patch, so I believe these errors are unrelated. Pull Request resolved: https://github.com/facebookresearch/pytorch3d/pull/230 Reviewed By: gkioxari Differential Revision: D23278126 Pulled By: nikhilaravi fbshipit-source-id: c06e891bc46de8222325ee7b37aa43cde44648e8

view details

Jeremy Reizenstein

commit sha e25ccab3d9e4d5c8877627ab7d95dd3f77c9a993

align_corners and padding for TexturesUV Summary: Allow, and make default, align_corners=True for texture maps. Allow changing the padding_mode and set the default to be "border" which produces more logical results. Some new documentation. The previous behavior corresponds to padding_mode="zeros" and align_corners=False. Reviewed By: gkioxari Differential Revision: D23268775 fbshipit-source-id: 58d6229baa591baa69705bcf97471c80ba3651de

view details

Jeremy Reizenstein

commit sha 909dc835050f84fc739fe75fd02884f305195afb

amalgamate meshes with texture into a single scene Summary: Add a join_scene method to all the textures to allow the join_mesh function to include textures. Rename the join_mesh function to join_meshes_as_scene. For TexturesAtlas, we now interpolate if the user attempts to have the resolution vary across the batch. This doesn't look great if the resolution is already very low. For TexturesUV, a rectangle packing function is required, this does something simple. Reviewed By: gkioxari Differential Revision: D23188773 fbshipit-source-id: c013db061a04076e13e90ccc168a7913e933a9c5

view details

Jeremy Reizenstein

commit sha 32484500befa825b3def2cdc15cb12eb2251f681

tutorial fixes from #336. Wheels with cuda10.1. Summary: Add a document to explain how to run the tutorials. Fix API of TexturesVertex in fit_textured_mesh. Prepare cuda 10.1 wheels (not 10.2) for linux to be available on pypi - this matches what colab has. Change the tutorials to use these new wheels. Reviewed By: gkioxari Differential Revision: D23324479 fbshipit-source-id: 60e92a3f46a2d878f811b7703638f8d1dae143d9

view details

Jeremy Reizenstein

commit sha 9aeb88b48e3c98e9229f28933cc9e1c2700a593f

tutorial fixes. validate #cameras Summary: Update the installation cells to import torch. Replace use of deprecated Textures in deform tutorial. Correct the deform tutorial's idea of its own name. Fix typo in batched part of render_textured_meshes which meant only one mesh was being rendered with a batch of cameras. Add an error check in the rasterizer to make the error friendly from such a mistake elsewhere. Reviewed By: gkioxari Differential Revision: D23345462 fbshipit-source-id: 1d5bd25db052f7ef687b7168d7aee5cc4dce8952

view details

Nikhila Ravi

commit sha c25fd836942c42101c7519c5124ebbb74d450bf8

small website updates Summary: Small fixes to website rendering Reviewed By: jcjohnson Differential Revision: D23281746 fbshipit-source-id: c9dc8edd5e52f39d4e0e19f10ecc7e035b39feda

view details

Steve Branson

commit sha 46c0e834616f6950d0ed5a9826a49744d8e83f7e

Change default settings of clip_barycentric_coords Summary: When raster_settings.clip_barycentric_coords is unspecified, clip_barycentric_coords is turned on if blur_radius is greater than 0. This matches the behavior prior to D21705503 (https://github.com/facebookresearch/pytorch3d/commit/cc70950f4064e3feeb55281b829aa55aa4a7e942). Reviewed By: gkioxari Differential Revision: D23375257 fbshipit-source-id: 4b87588aabb69d4d835d4dcceb11153628121d30

view details

Jeremy Reizenstein

commit sha 7414904c904f43b231761939fb60fdcbb34ff586

Installation updates for next release. Summary: We've made some build updates. Reviewed By: jcjohnson Differential Revision: D23244111 fbshipit-source-id: dba6e5dc3a9d7dcb2b62094f7f7b8b5abca190d9

view details

Jeremy Reizenstein

commit sha 701bbef4f34001c47ec2631ffd20d30d8fc241ea

version number for version 0.2.5 Reviewed By: jcjohnson Differential Revision: D23244140 fbshipit-source-id: 8f95c817637162350b61b1b0fc3d2de66087250e

view details

push time in 16 days

push eventyifita/pytorch3d

Nikhila Ravi

commit sha 8e9ff15faff8acf6bf5c13e19a7e2790e61af470

datasets.md (documentation for datasets) Summary: documentation for datasets Reviewed By: bottler, gkioxari Differential Revision: D22992266 fbshipit-source-id: 44aaa8227af04c1baa5ea6c31ba131dea2b9675b

view details

ywang

commit sha 8baba46b481e8eab561a9abc21242a78dd124906

Merge branch 'master' of https://github.com/facebookresearch/pytorch3d

view details

push time in a month

push eventyifita/pytorch3d

Jeremy Reizenstein

commit sha 4872a2c4def6b773482766beb0a1729a0724c809

wheels with cuda Summary: The main pytorch wheels on PyPI support CUDA 10.2. Here we make pytorch3d's wheels do the same, instead of being cpu only. This should ultimately make life easier in colab. Also a little script to count builds, which can be useful for nightly jobs. Reviewed By: gkioxari Differential Revision: D22924321 fbshipit-source-id: d6cea9bfbab49bcb0080f65608066c553ea8bb4d

view details

Jeremy Reizenstein

commit sha 7944d24d4872bdb01b821450840049e28d0ce12b

gather_scatter on CPU Summary: CPU implementation of the graph convolution op. Reviewed By: nikhilaravi, gkioxari Differential Revision: D21384361 fbshipit-source-id: bc96730e9727bb9aa1b0a232dcb82f0c0d12fe6b

view details

Jeremy Reizenstein

commit sha 5d9444307f5ccff5fdc15b7b1ed3eaf526c0917e

fix graph_conv test Summary: The recently added part of a test was assuming that the random gpu was gpu 0. Reviewed By: nikhilaravi Differential Revision: D22948397 fbshipit-source-id: 88107e19fc3118e763f95be43a614941176a08f9

view details

Jeremy Reizenstein

commit sha c122ccb13c51912273f93fba8fe45fb0abf051c3

No more cuda 10.0 builds Summary: Latest pytorch doesn't support cuda 10.0 and now its pytorch/conda-cuda docker image doesn't include it either. Here we remove the pytorch3d builds which use cuda 10.0. Reviewed By: nikhilaravi Differential Revision: D22999392 fbshipit-source-id: c834477fc7c812c2d0594dccd9e9471e33a4ec5e

view details

Luya Gao

commit sha 326e4ccb5bc3de3bdee97cc5907e2e2c75bdcc73

Return R2N2 R,T,K Summary: Return rotation, translation and intrinsic matrices necessary to reproduce R2N2's own renderings. Reviewed By: nikhilaravi Differential Revision: D22462520 fbshipit-source-id: 46a3859743ebc43c7a24f75827d2be3adf3f486b

view details

Luya Gao

commit sha 63ba74f1a8d5aa73d36999f5b5a7bf2af0fd8066

Return R2N2 voxel coordinates Summary: Return R2N2's voxel coordinates. Reviewed By: nikhilaravi Differential Revision: D22462530 fbshipit-source-id: a995cfa0957b2561eb3b0f4591cb1db42170bc68

view details

Patrick Labatut

commit sha 6d76336501918e29b9cc15b5b4fe6d0f18818d67

Extract more reusable I/O functions Summary: Continue extracting reusable I/O functions to a separate utils module (and remove duplication). Reviewed By: nikhilaravi Differential Revision: D20720433 fbshipit-source-id: e82b19560a5dc8a506c4c4d098da69c202790c4f

view details

ywang

commit sha 7666b63e883751b11c9c631fe5e4b2ba8e2be28f

Merge branch 'master' of https://github.com/facebookresearch/pytorch3d

view details

push time in a month

push eventyifita/differentiable_volumetric_rendering

ywang

commit sha 67286194dbc8075edc2a8f9303007fad740a4d7f

depth_func accept sdf decoder

view details

ywang

commit sha 110cbde0b727afc3d20adae90c338a375e3e7672

change decoder output dim

view details

ywang

commit sha 3d59d95e9e2a3e2a5544ac4587df54993a0bd057

gitignore

view details

push time in 2 months

issue commentyifita/deep_cage

Compiling error

Please try to install with 1.4.0, I'll correct the environment file in pytorch_points. Thanks for the issue!

wanglei1006

comment created time in 2 months

push eventyifita/yifita.github.io

ywang

commit sha fc55ac088af4437c72dbeecdea0c26e4a4e07677

rebuilding site Di Jul 14 11:24:36 CEST 2020

view details

push time in 2 months

push eventyifita/deep_cage

ywang

commit sha e0db2175128c7aaf3add513f3eca1afe90b8e8e5

use face correspondence instead of vertex

view details

push time in 2 months

startedvsitzmann/siren

started time in 3 months

issue commentyifita/deep_cage

Compiling error

which pytorch version are you using? This looks like a version incompatibility issue.

wanglei1006

comment created time in 3 months

issue commentyifita/deep_cage

denormalize the deform

Thanks, but I think i'll not address this feature any time soon :)

hjwdzh

comment created time in 3 months

issue closedyifita/deep_cage

denormalize the deform

Currently, the output of the cage_deform is the deformation that fits the normalized target shape. I think it might be a good idea to instead provide a denormalization :)

closed time in 3 months

hjwdzh

issue commentyifita/deep_cage

Training settings about SURREAL

Hi @zhangyujie9999 , thanks for trying out our code. Did you check out the code I recently uploaded? See the new Readme.md page for instructions for SURREAL. https://github.com/yifita/deep_cage#humanoid-deformations

zhangyujie9999

comment created time in 3 months

issue commentyifita/deep_cage

Training setting

Hi Steve, I uploaded the instructions to train SURREAL data. Feel free to test out and let me know if you have questions.

Steve-Tod

comment created time in 3 months

issue closedyifita/deep_cage

`faiss_knn, group_knn` not implemented in `pytorch_points.network.operations.py`

Hi, Yifan,

Thanks for sharing this great work. However, when I finished the setup, and run the demo code, I found the following error,

ImportError: cannot import name 'faiss_knn' from 'pytorch_points.network.operations' (xxx/deep_cage/pytorch_points/pytorch_points/network/operations.py)

I have doublechecked your pytorch_points project. There are no faiss_knn or group_knn function in the operations.py script.

closed time in 3 months

KnightOfTheMoonlight

issue commentyifita/deep_cage

`faiss_knn, group_knn` not implemented in `pytorch_points.network.operations.py`

Hi @KnightOfTheMoonlight thanks for using the code. You need to use the pytorch_points version reference in the submodule, which contains these two functions.

To do that, make sure that you clone the neural cage repo with --recursive option, i.e. git clone --recursive https://github.com/yifita/deep_cage.git

KnightOfTheMoonlight

comment created time in 3 months

more