Instructions To Reproduce the Issue:

commit sha 9a0d1cc9d0a7113e66d23cfad0183c1dec674c9f

Refactored Query pages, removed unnecessary code

commit sha 7f9bc9c79f09f2adb7dc40cdd9b50d7d277ad3af

Merge branch 'refactorQuery' into 'master' Refactored Query pages, removed unnecessary code See merge request l3pilot-sp5/pc_db_ui!58

push time in a month

push eventeduardohenriquearnold/l3p-ui

Sajjad

commit sha 98aaef27bf50940e754da1aea09e0728d5a1ee3d

Fixed csv format issue regarding headings and extra comma after each page of query

Sajjad Mozaffari

commit sha 3b0acb5cad4937cd1e44c65414ef11bbc7439bc9

Merge branch '49-formatting-of-csv-files-for-export' into 'master' Resolve "Formatting of csv files for export" Closes #49 See merge request l3pilot-sp5/pc_db_ui!56

commit sha 135296bb9f7994e4d51c9bef57ded21bb686d0c9

Improved LoginPage aesthetic

commit sha 98f5d07381e2ffe4f7982199b7f489aa5b2ca066

Merge branch 'newlogin' into 'master' Improved LoginPage aesthetic See merge request l3pilot-sp5/pc_db_ui!57

commit sha 1f6094db6422e9866f13be4e91a482e7345d562d

Merging github change that wasn't commited to Gitlab

push time in a month

push eventeduardohenriquearnold/l3p-ui

Sajjad

commit sha 87dc170a17b55dcfa9db5e54e0ef8bd8a74d6332

Download now skip empty measurements,file extentions are now corrected (csv)

commit sha 5c8cbda35f17b7326f949c54368d403ce55e69da

Added Documentation link to nav-bar

commit sha 8e1af3f0379c5b733ace70a816b5be0680861ac1

Merge branch 'doc-menu' into 'master' Doc menu See merge request l3pilot-sp5/pc_db_ui!55

push time in a month

startedmitsuba-renderer/mitsuba2

started time in a month

MemberEvent

push eventeduardohenriquearnold/eduardohenriquearnold.github.io

commit sha 0b9580780a508256969c13f9c21dc1cb763e2c4b

Build Sat 25 Jul 17:54:55 BST 2020

push time in 2 months

push eventeduardohenriquearnold/eduardohenriquearnold.github.io

commit sha 7cdc5fb754ef784d1f20c849a9658c8905e83e29

Updated timeseries forecasting blog post and notebook

push time in 2 months

push eventeduardohenriquearnold/eduardohenriquearnold.github.io

commit sha 8b98e5dcbd43a40a45660316bd516938ee275ce8

Added bayesian linear regression post (from March2019)

push time in 2 months

push eventeduardohenriquearnold/eduardohenriquearnold.github.io

commit sha 074fd8f331efb273145c83b0701080eb4f702cf6

Build Thu 23 Jul 20:41:50 BST 2020

push time in 2 months

push eventeduardohenriquearnold/eduardohenriquearnold.github.io

commit sha 1053fd8376603b30c4c53bc6f58515e252676bf7

Added IET talk slides

commit sha 28b96d488db321941e24d1b9ff8fef25ba4ff1ad

Created a new post on timeseries forecasting

push time in 2 months

push eventeduardohenriquearnold/eduardohenriquearnold.github.io

commit sha a69f2593f7784204c7b371a8f66c653cf9e501fe

Build Thu 23 Jul 16:43:33 BST 2020

push time in 2 months

push eventeduardohenriquearnold/eduardohenriquearnold.github.io

commit sha 7106669a0159eb4786f77527de10bb9515d82aac

Build Mon 20 Jul 12:51:34 BST 2020

push time in 2 months

push eventeduardohenriquearnold/eduardohenriquearnold.github.io

commit sha 5dffbcffbcf31abda523424e903bd0f03ea3c805

Added earnold.me domain name to config.toml and the static CNAME required by GH pages

commit sha 01766b696122066402541bb4c5e7234f20388ef6

Updated publications

commit sha 3ab8b7d5f4c1c97ee6299dc740724041defd5f5c

Updated talks

push time in 2 months

push eventeduardohenriquearnold/eduardohenriquearnold.github.io

commit sha 5f76e0fcc2022ee0ea2423ed9c64f87e662f33e4

Build Mon 20 Jul 12:28:54 BST 2020

push time in 2 months

push eventeduardohenriquearnold/eduardohenriquearnold.github.io

commit sha e524898ac6403e5aa20599cc8a5f6bfbb3ceccc2

Build Tue 14 Jul 11:02:32 BST 2020

push time in 2 months

push eventeduardohenriquearnold/eduardohenriquearnold.github.io

commit sha 1c1ea8fa0230d15cc07b3bbd715ded9a3ce291ef

Create CNAME

so3_log_map unexpected behaviour

push time in 2 months

startederdos-project/pylot

started time in 2 months

issue closedfacebookresearch/pytorch3d

Unexpected output of so3_log_map. so3_log_map and so3_exponential_map are inverse functions. However composing them for some rotation matrices do not result in the original rotation matrix R. Please note the code below.

Instructions To Reproduce the Issue:

import torch
from pytorch3d.transforms import so3_exponential_map, so3_log_map
from pytorch3d.renderer import look_at_view_transform

R,T = look_at_view_transform(1, elev=50, azim=0)
print(R)

log_R = so3_log_map(R, 1e-8)
R_hat = so3_exponential_map(log_R)
print(R_hat)

The output is

tensor([[[-1.0000,  0.0000,  0.0000],
         [ 0.0000,  0.6428, -0.7660],
         [-0.0000, -0.7660, -0.6428]]])
tensor([[[1., 0., 0.],
         [0., 1., 0.],
         [0., 0., 1.]]])

The matrices do not match, as one would expect. I cannot tell why, but I have noticed that this happens for any elevation angle <90 when azim=0.

closed time in 2 months

eduardohenriquearnold

issue commentfacebookresearch/pytorch3d

so3_log_map unexpected behaviour

Hi. Sorry, I should have realised this earlier. Thanks for the great explanation though.

eduardohenriquearnold

comment created time in 2 months

push eventeduardohenriquearnold/eduardohenriquearnold.github.io

commit sha 400ad7435a0178125231a3cbfde686a573296b9d

Removed old academic theme version

commit sha fdc3591e1320eb9da699b74c48883ff6676215cd

Updated to new Academic theme

commit sha 9bd374a00694e51c4f5111eef08b25fa6b49df73

Added hugo academic theme as a submodule

push time in 2 months

push eventeduardohenriquearnold/eduardohenriquearnold.github.io

commit sha 7c114ca11416286be43b4bcee17b0364296575d3

Build Mon 13 Jul 15:58:29 BST 2020

push time in 2 months

push eventeduardohenriquearnold/eduardohenriquearnold.github.io

commit sha 7cbb6869bb004042571c168576e9c928fb960929

Updated about info with Warwick information. Updated portrait pic

push time in 2 months

push eventeduardohenriquearnold/eduardohenriquearnold.github.io

commit sha 2a9c9fed1965327f8529161400ca29d041f0b4d6

Build Sun Jul 12 20:45:53 BST 2020

so3_log_map unexpected behaviour

push time in 2 months

PublicEvent

issue openedfacebookresearch/pytorch3d

Instructions To Reproduce the Issue:

import torch
from pytorch3d.transforms import so3_exponential_map, so3_log_map
from pytorch3d.renderer import look_at_view_transform

R,T = look_at_view_transform(1, elev=50, azim=0)
print(R)

log_R = so3_log_map(R, 1e-8)
R_hat = so3_exponential_map(log_R)
print(R_hat)

The output is

tensor([[[-1.0000,  0.0000,  0.0000],
         [ 0.0000,  0.6428, -0.7660],
         [-0.0000, -0.7660, -0.6428]]])
tensor([[[1., 0., 0.],
         [0., 1., 0.],
         [0., 0., 1.]]])

The matrices do not match, as one would expect. I cannot tell why, but I have noticed that this happens for any elevation angle <90 when azim=0.

created time in 2 months

issue closedfacebookresearch/pytorch3d

When using knn_points on CUDA with large K (e.g. 4000) the code fails with the error

THCudaCheck FAIL file=/opt/conda/conda-bld/pytorch_1587428111115/work/aten/src/THC/THCReduceAll.cuh line=327 error=6 : the launch timed out and was terminated
Traceback (most recent call last):
  File "testknn.py", line 11, in <module>
    d = knn_points(g,p,None,None,K).dists
  File "/home/eduardo/Documents/Experiments/pytorch3d/pytorch3d/ops/knn.py", line 161, in knn_points
    p1_dists, p1_idx = _knn_points.apply(p1, p2, lengths1, lengths2, K, version)
  File "/home/eduardo/Documents/Experiments/pytorch3d/pytorch3d/ops/knn.py", line 56, in forward
    if lengths2.min() < K:
RuntimeError: cuda runtime error (6) : the launch timed out and was terminated at /opt/conda/conda-bld/pytorch_1587428111115/work/aten/src/THC/THCReduceAll.cuh:327

Tried all different versions of KNN (0 to 3) and all of them result in the same error. Using a smaller K (e.g. 2000) works just fine. If you change the device to cpu it works just fine. It does not seem to be a memory problem as the GPU memory is way below the maximum RAM of 8Gb.

Using pytorch3d compiled from source on commit 7f1e63aed1252ba8145d4a.

To Reproduce the Issue:

import torch
from pytorch3d.ops import knn_points

K = 4000
device = torch.device('cuda')
p = torch.rand(16, 200000, 3).to(device)
g = torch.rand(16, 8, 3).to(device)
d = knn_points(g, p, None, None, K).dists
print(d.shape)

closed time in 2 months

eduardohenriquearnold

issue commentfacebookresearch/pytorch3d

Thanks, I believe this solves my problem!

eduardohenriquearnold

comment created time in 2 months

startedfacebookresearch/detr

started time in 2 months

issue commentfacebookresearch/pytorch3d

Thanks for the prompt responses.

The PyTorch3D KNN is optimized for K < 32 and D < 4 (where D is the feature size). What is your use case for needing to use K = 4000?

A very large K doesn't seem to make much sense but I will try and explain my case. I'm optimising the sensor pose such that each object that I want to observe has a very large number of points on their surfaces - more accurately, I want to maximise the average number of points over all objects' surfaces. Since the average number of points on objects' surface is not differentiable itself (the counting operation is not differentiable), I am optimising the L2 distance between the centre of the objects and the K nearest points in their neighbourhood. However I need K >> 32, e.g. 1k or 4k. The KNN with large K seems to work fine on the CPU, but if I move the whole pipeline to the CPU my execution time increases drastically. Perhaps there is a simpler loss function to maximise the number of points on the surface of objects? I tried using the Chamfer loss between all points and the centre of objects, but the sensors tend to just collapse into some of the objects, thus I decided to limit the distance to K points only.

Maybe the calculation is just taking a long time in your case. If you are using the GPU for display as well as computation, the GPU can automatically kill long-running functions. See e.g. here. Can you split your calculation along the batch dimension so that each one will be quicker?

Thanks for the suggestion. Unfortunately I cannot disable my X server since I need to generate a visualisation of the optimisation process. Splitting the calculation to the batch dim did help improve to improve the speed when K=1000, but still resulted in the kernel being killed for K=4000.

eduardohenriquearnold

comment created time in 2 months

issue openedfacebookresearch/pytorch3d

Fix look_at_view_transform when object location is not (0,0,0)

When using knn_points on CUDA with large K (e.g. 4000) the code fails with the error

THCudaCheck FAIL file=/opt/conda/conda-bld/pytorch_1587428111115/work/aten/src/THC/THCReduceAll.cuh line=327 error=6 : the launch timed out and was terminated
Traceback (most recent call last):
  File "testknn.py", line 11, in <module>
    d = knn_points(g,p,None,None,K).dists
  File "/home/eduardo/Documents/Experiments/pytorch3d/pytorch3d/ops/knn.py", line 161, in knn_points
    p1_dists, p1_idx = _knn_points.apply(p1, p2, lengths1, lengths2, K, version)
  File "/home/eduardo/Documents/Experiments/pytorch3d/pytorch3d/ops/knn.py", line 56, in forward
    if lengths2.min() < K:
RuntimeError: cuda runtime error (6) : the launch timed out and was terminated at /opt/conda/conda-bld/pytorch_1587428111115/work/aten/src/THC/THCReduceAll.cuh:327

Tried all different versions of KNN (0 to 3) and all of them result in the same error. If you change the device to cpu it works just fine. It does not seem to be a memory problem as the GPU memory is way below the maximum RAM of 8Gb.

Using pytorch3d compiled from source on commit 7f1e63aed1252ba8145d4a.

To Reproduce the Issue:

import torch
from pytorch3d.ops import knn_points

K = 4000
device = torch.device('cuda')
p = torch.rand(16, 200000, 3).to(device)
g = torch.rand(16, 8, 3).to(device)
d = knn_points(g, p, None, None, K).dists
print(d.shape)

created time in 3 months

pull request commentfacebookresearch/pytorch3d

Just added a unit test for look_at_view_transform when at != (0,0,0).

eduardohenriquearnold

comment created time in 3 months

push eventeduardohenriquearnold/pytorch3d

commit sha 3759186a8ea30cf889a4b4a986fe05b68d5c5c27

Add unit test for look_at_view_transform when 'at' is not (0,0,0)