profile
viewpoint

jcjohnson/neural-style 17354

Torch implementation of neural style algorithm

jcjohnson/fast-neural-style 3452

Feedforward style transfer

jcjohnson/pytorch-examples 3429

Simple examples to introduce PyTorch

jcjohnson/cnn-benchmarks 2188

Benchmarks for popular CNN models

jcjohnson/densecap 1356

Dense image captioning in Torch

google/sg2im 1063

Code for "Image Generation from Scene Graphs", Johnson et al, CVPR 2018

jcjohnson/cnn-vis 493

Use CNNs to generate images

jcjohnson/simple-amt 234

A microframework for working with Amazon's Mechanical Turk

jcjohnson/neural-animation 13

Implementing neural art on video

startedmichaelrzhang/lookahead

started time in 3 days

startedfacebookresearch/WSL-Images

started time in 7 days

push eventjcjohnson/pytorch-vgg

Justin Johnson

commit sha 121d07b320b606ba2052efe2dca1398db46f186d

Update README.md

view details

push time in 9 days

issue commentfacebookresearch/pytorch3d

Optimize Object Pose with a PhongShader instead of SilhouetteShader

Can you post a bit of code that reproduces the problem? That will make it easier for us to help.

filipetrocadoferreira

comment created time in 14 days

issue commentfacebookresearch/pytorch3d

import error

Does it work if you import pytorch3d before importing _C?

import pytorch3d
from pytorch3d import _C
amiltonwong

comment created time in 16 days

startedfacebookresearch/faiss

started time in 21 days

startedfacebookresearch/meshrcnn

started time in a month

startedfacebookresearch/pytorch3d

started time in a month

push eventjcjohnson/website

Justin Johnson

commit sha 0058473661f170c53bdd236f686eedb5fe680b45

Add code for HiDDeN

view details

push time in a month

push eventjcjohnson/website

Justin Johnson

commit sha 629e39ce96ce075088794978d4de42169c62a5c6

Remove stray link

view details

push time in a month

push eventjcjohnson/website

Justin Johnson

commit sha fc9aff6e32bf55f8df9cf0de00a22596c3bf6638

Redirect stanford page to UMich

view details

push time in a month

push eventjcjohnson/website

Justin Johnson

commit sha df2b35e5d1cccc6292c947150d7b197a3d356530

Long overdue update

view details

push time in a month

create barncheecs442/utils

branch : master

created branch time in a month

created repositoryeecs442/utils

Student-facing utilities for EECS 442

created time in a month

pull request commentjcjohnson/simple-amt

Related to #29. Setting default region to us-east-1

This looks good to me, thanks!

shubhamagarwal92

comment created time in a month

push eventjcjohnson/simple-amt

Shubham Agarwal

commit sha 3a4422641c4aa7933c4c62511b78acfdff7a1ef1

Related to #29. Setting default region to us-east-1

view details

Justin Johnson

commit sha 3001e1a663c2bc32706cef8568c7c4f0d7ced640

Merge pull request #30 from shubhamagarwal92/master Related to #29. Setting default region to us-east-1

view details

push time in a month

PR merged jcjohnson/simple-amt

Related to #29. Setting default region to us-east-1

boto3 client requires a region to make a connection. if you have a default region in your ~/.aws/config other than us-east-1, it throws an error. Since Mturk endpoint is by default only in us-east-1, there is no point of asking users to provide it. See #29

+11 -2

0 comment

2 changed files

shubhamagarwal92

pr closed time in a month

issue commentgoogle/sg2im

Question about Inception Scores

Inception scores are extremely sensitive to the exact model you use to compute them.

It looks like the code you link is using the torchvisvion inception_v3 model; however most people who report inception scores (including my paper) typically use the Inception v1 model from Google. I used the original OpenAI code to compute inception scores:

https://github.com/openai/improved-gan/blob/master/inception_score/model.py

zx1301

comment created time in 3 months

issue commentfacebookresearch/clevr-dataset-gen

Blender compatible version

I haven't tested any of this code with Blender 2.8, so I don't know if it will work. You should stick with 2.7.

The last value in pixel coords gives the z coordinate of the object center; I don't recall exactly what the units are, but you can use it to compare the relative depths of objects.

AvivSham

comment created time in 3 months

startedNVIDIAGameWorks/kaolin

started time in 3 months

startedfacebookresearch/nds

started time in 4 months

startedfacebookresearch/pycls

started time in 4 months

delete branch deepvision-class/starter-code

delete branch : A4

delete time in 4 months

push eventdeepvision-class/starter-code

Luowei Zhou

commit sha 9a172a78b370811b972ad351d1ae461bd24074bd

added decode_captions and attention_visualizer

view details

Justin Johnson

commit sha e8d9fe711870a39796a2f8ad95538e57942d756f

Merge pull request #3 from deepvision-class/A4 added decode_captions and attention_visualizer

view details

push time in 4 months

issue closedjcjohnson/pytorch-examples

Why does `nn` module need a higher learning rate?

Thanks a lot for your amazing examples!

I find that in the examples using nn module the learning_rate is 1e-4, while without the nn module the learning_rate is 1e-6.

I did not figure out the reason why those examples using nn require larger learning_rate🙁

closed time in 4 months

Liwink

issue closedjcjohnson/pytorch-examples

pytorch overhead

Hey @jcjohnson, sorry for opening an issue here, didn't know how to reach You without e-mailing on Your Stanford inbox.

I haven't used torch or Lua, but I remember some of my friends talking about your implementation of char-rnn in Lua. They said it was super fast.

I'm wondering if it is possible to do something like that in PyTorch? Or the speed was thanks to Lua's JIT compiler, and Python interpreter will simply incur too much overhead? In general, do You think PyTorch is suitable for applications with lots of small computations (char-lvl, pixel-lvl stuff)?

closed time in 4 months

elanmart

issue closedjcjohnson/pytorch-examples

Adding some layers

closed time in 4 months

MitraTj

issue closedjcjohnson/pytorch-examples

Issue on Windows

I try to install pytorch and after days of trying Im here with a big, big problem. I read a lot of articles of "how to install pytorch" I try to install with pip install but dont work for me and after I install it with Anaconda, but in anaconda is pytorch install, when I type: conda list, he is there like this form: pytorch 1.0.1 py3.7_cuda100_cudnn7_1 pytorch, I have python 3.7, when I run a code with import torch this show me a message like this:

problem

And when i try to import torch in python 3.7:

fdzzddz

Pip install error:

dsfdasad

How to pass this errors? Please Help, thx.

closed time in 4 months

FabyOfficial

issue commentjcjohnson/pytorch-examples

Issue on Windows

This is an installation issue with Pytorch, and has nothing to do with the content of this repo.

FabyOfficial

comment created time in 4 months

issue commentgoogle/sg2im

Vocab mismatch between checkpoint and paper

My reply times are usually bimodal: either I respond right away or it will fall out of my inbox and be forgotten forever!

aluo-x

comment created time in 4 months

issue commentgoogle/sg2im

Vocab mismatch between checkpoint and paper

In some of my earlier experiments I tried some additional relationships on COCO, defined as follows (replaces https://github.com/google/sg2im/blob/master/sg2im/data/coco.py#L337)

      if sx0 < ox0 and sx1 > ox1 and sy0 < oy0 and sy1 > oy1:
        p = 'surrounding'
      elif sx0 > ox0 and sx1 < ox1 and sy0 > oy0 and sy1 < oy1:
        p = 'inside'
      elif theta >= 3 * math.pi / 4 or theta <= -3 * math.pi / 4:
        p = 'right touching' if touching else 'left of'
      elif -3 * math.pi / 4 <= theta < -math.pi / 4:
        p = 'bottom touching' if touching else 'above'
      elif -math.pi / 4 <= theta < math.pi / 4:
        p = 'left touching' if touching else 'right of'
      elif math.pi / 4 <= theta < 3 * math.pi / 4:
        p = 'top touching' if touching else 'below'
      p = self.vocab['pred_name_to_idx'][p]
      triples.append([s, p, o])

However in the final models I didn't end up using these relationships. They are still present in the vocab of the pretrained models, but these relationships were not used at all during training and the embeddings associated with these relationships will be random. Thus if you try to pass a scene graph with one of these "touching" relationships, you will probably get a garbage output from the model.

aluo-x

comment created time in 4 months

startedfacebookresearch/detectron2

started time in 4 months

delete branch deepvision-class/starter-code

delete branch : A3

delete time in 5 months

PR merged deepvision-class/starter-code

A3
+391 -3

0 comment

4 changed files

jcjohnson

pr closed time in 5 months

push eventdeepvision-class/starter-code

kibok

commit sha 6f5d0ff8f3443a62ffe540d25abefa102ff33185

Update for A3

view details

kibok

commit sha 2eb45cc07b4e9817a0d8784facd2f2ce208fcd57

Add extract_drive_file_id to import_colab

view details

kibok

commit sha 1d060b9082ae5c7f671941256867cd50d07b537e

Add time limit in Solver.train

view details

Justin Johnson

commit sha d45bf7b3d790d9d3549b97559a56a2e26b015f26

Merge pull request #2 from deepvision-class/A3 A3

view details

push time in 5 months

PR opened deepvision-class/starter-code

A3
+391 -3

0 comment

4 changed files

pr created time in 5 months

push eventjcjohnson/simple-amt

Kaiyu Yang

commit sha d70b27571244ddb406054c544f4c155b6dba6536

upgrade to boto 3

view details

Kaiyu Yang

commit sha d6bfafc3979690a5d5da4b2a2b19868162a12db9

fix minor issues

view details

Kaiyu Yang

commit sha ad93964af31e0294d8f87c0d45536e041db60e79

Update simpleamt.py

view details

Justin Johnson

commit sha df32687b686d639cc49b6df225831ff56adf1030

Merge pull request #28 from yangky11/master Update the Code to Use boto 3

view details

push time in 5 months

pull request commentjcjohnson/simple-amt

Update the Code to Use boto 3

Thanks!

yangky11

comment created time in 5 months

pull request commentjcjohnson/simple-amt

Update the Code to Use boto 3

Is this ready to merge now?

yangky11

comment created time in 5 months

PR merged jcjohnson/simple-amt

Update the Code to Use boto 3

Related to this issue.

Not finished yet, but launch_hits.py has been fixed.

+174 -167

5 comments

21 changed files

yangky11

pr closed time in 5 months

issue commentgoogle/sg2im

Image quality

The model is an example of a generative adversarial network; the mathematical formulation requires noise as input to the system so that the output can model a full probability distribution.

In this case I suspect that models suffer from mode collapse, and thus don't actually make much use of the noise.

Since this project is licensed under Apache-2.0 you can use this project for commercial purposes (subject to the constraints of the license) but you should have some realistic expectation about the results you will get since the project wasn't intended for business use.

SandeepBhutani

comment created time in 5 months

issue closedgoogle/sg2im

Is this project still being worked upon?

I can see this project was last updated 11 months ago. Wondering if someone is already working on this?
Was planning to use to it generate and altering images while someone describes them.

closed time in 5 months

SandeepBhutani

issue commentgoogle/sg2im

Is this project still being worked upon?

This is the code for a research paper published at CVPR 2018. I'm not actively working on it at this time.

SandeepBhutani

comment created time in 5 months

issue commentgoogle/sg2im

Image quality

What do you mean by "noise is being added in code"? How else would you expect noise to be added?

This was a research project and not intended for business or production use.

SandeepBhutani

comment created time in 5 months

delete branch deepvision-class/starter-code

delete branch : assignment2

delete time in 5 months

push eventdeepvision-class/starter-code

Yunseok Jang

commit sha 4cf5a4bddc8edc19fd2942deabadc689f555eb5a

add two additional helper functions

view details

Justin Johnson

commit sha b51cf56615478f81509c43a357fc86a4af49eb71

Merge pull request #1 from deepvision-class/assignment2 add two additional helper functions

view details

push time in 5 months

push eventdeepvision-class/tools

朱奕安

commit sha bafa0257f1dcfab6faf55fa782bbf68898bf903f

fix: arg key of add_argument() missing get error msg "ValueError: invalid option string 'assignment': must start with a character '-'" otherwise

view details

Justin Johnson

commit sha be6e393c4b9f8c8f458278b8230d3c2be237e9ce

Merge pull request #1 from Angus-ZHU/master fix: arg key of add_argument() missing

view details

push time in 5 months

PR merged deepvision-class/tools

fix: arg key of add_argument() missing

get error msg "ValueError: invalid option string 'assignment': must start with a character '-'" otherwise

+1 -1

0 comment

1 changed file

Angus-ZHU

pr closed time in 5 months

push eventdeepvision-class/tools

Justin Johnson

commit sha 6bbe11a6fe0515a5a87d94f1b7854009a82b9136

Update README.md

view details

push time in 5 months

push eventdeepvision-class/tools

Justin Johnson

commit sha 0523c929d3f637ec2122032af59f20994332ddce

Create README.md

view details

push time in 5 months

create barnchdeepvision-class/tools

branch : master

created branch time in 5 months

created repositorydeepvision-class/tools

Student-facing utility scripts

created time in 5 months

more