Comments (8)
It's the CUDA library for bilinear sampling, from this repository:
https://github.com/qassemoquab/stnbhwd
You can install it like this:
cd toflow/src/stnbhwd
luarocks make
Please let us know if it doesn't work!
Thanks!
from toflow.
When I installed it, it installed successfully, but I can't find the lib named 'libcustn' and meet the following info. The error above still exists
`-- Found Torch7 in /data/guo/torch/install
-- Compiling with OpenMP support
-- Could NOT find CUDA (missing: CUDA_INCLUDE_DIRS CUDA_CUDART_LIBRARY) (found suitable version "8.0", minimum required is "5.5")
-- Configuring done
-- Generating done
-- Build files have been written to: /data/guo/toflow-master/src/stnbhwd/build
[100%] Built target stn
from toflow.
Let me try to help you on this.
Would you like to give me your system configuration? Are you using CUDA / GPU or not?
Could you run require 'stn'
by itself and see whether it works?
Make sure you run everything with flag -nocuda
if you are not using CUDA/GPU.
For example th demo.lua -mode interp -inpath ../data/example/low_frame_rate -nocuda
from toflow.
I run require 'stn'
and th demo.lua -mode interp -inpath ../data/example/low_frame_rate -nocuda
, but they all throw out the same error.
/data/guo/torch/install/share/lua/5.1/trepl/init.lua:389: /data/guo/torch/install/share/lua/5.1/trepl/init.lua:389: module 'libcustn' not found:No LuaRocks module found for libcustn no field package.preload['libcustn'] no file '/root/.luarocks/share/lua/5.1/libcustn.lua' no file '/root/.luarocks/share/lua/5.1/libcustn/init.lua' no file '/data/guo/torch/install/share/lua/5.1/libcustn.lua' no file '/data/guo/torch/install/share/lua/5.1/libcustn/init.lua' no file '/data/ivenhyxu/stereo_match_dl/distro/install/share/lua/5.1/libcustn.lua' no file '/data/ivenhyxu/stereo_match_dl/distro/install/share/lua/5.1/libcustn/init.lua' no file './libcustn.lua' no file '/data/ivenhyxu/stereo_match_dl/distro/install/share/luajit-2.1.0-beta1/libcustn.lua' no file '/usr/local/share/lua/5.1/libcustn.lua' no file '/usr/local/share/lua/5.1/libcustn/init.lua' no file '/root/.luarocks/lib/lua/5.1/libcustn.so' no file '/data/guo/torch/install/lib/lua/5.1/libcustn.so' no file '/data/guo/torch/install/lib/libcustn.so' no file '/data/ivenhyxu/stereo_match_dl/distro/install/lib/libcustn.so' no file '/data/ivenhyxu/stereo_match_dl/distro/install/lib/lua/5.1/libcustn.so' no file './libcustn.so' no file '/usr/local/lib/lua/5.1/libcustn.so' no file '/usr/local/lib/lua/5.1/loadall.so' stack traceback: [C]: in function 'error' /data/guo/torch/install/share/lua/5.1/trepl/init.lua:389: in function 'require' [string "_RESULT={require 'stn'}"]:1: in main chunk [C]: in function 'xpcall' /data/guo/torch/install/share/lua/5.1/trepl/init.lua:661: in function 'repl' ...rguo/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:204: in main chunk [C]: at 0x004064f0
Yes, I am using GPU.
The version of CUDA is 8.0.61
And the version of cudnn is 6.0.21
from toflow.
I tried this algorithm in another machine which system is Ubuntu with nocuda mode (This machine is no GPU and The system of the machine used before is CentOS ). It worked and the result was awesome, but it corrupted when I test deblock mode with my own picture(resolution:540x960 format:png). Are there some limits to the picture format?
==> initializing... input sequence: ../data/example/block result stored in: ../demo_output ==> loading... /home/guoyang/torch/install/bin/luajit: ./main/loader.lua:45: attempt to index local 'inputpath' (a nil value) stack traceback: ./main/loader.lua:45: in function 'get_file' demo.lua:75: in main chunk [C]: in function 'dofile' ...yang/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk [C]: at 0x00406670
from toflow.
Thanks!
We have only tested our project on Ubuntu, so we currently will not be able to help with problems on CentOS.
For the deblock problem, would you like to send us the command u execute and the file structure (specifically the location of the input image)? It looks more like a missing file error.
from toflow.
The command I used was one of your examples.
th demo.lua -mode deblock -nocuda -inpath ../data/example/block
I just replaced the pictures in the input directory.
When I added the number of pictures to 7, this error hadn't happen. In the function 'gen_path' of the file 'loader.lua', I find the value of the parameter 'num' is 7, is this the cause of this problem?
from toflow.
Yes! Our algorithm works with sequences of 7 frames in deblock task.
I will first close this issue (one issue per problem), feel free to open another issue if you need further help!
Thanks for your interest in your project!
from toflow.
Related Issues (20)
- Pre-training the flow estimation network HOT 7
- Resolution mismatching for interpolation HOT 2
- the noise level of the pretrained denoise model HOT 2
- vimeo-90k HOT 4
- Using ffmpeg to generate image sequence? HOT 1
- Looking forward to the training code. HOT 1
- divide vimeo90K dataset into training set and test set
- septuplet dataset 502 HOT 2
- Running toflow on consecutive frames HOT 1
- What is the number of parameters for this algorithm? HOT 1
- the results of sr model is not correct HOT 2
- add noise to input seems to have something wrong HOT 3
- Loading the model:
- Confusion on Image Transformation Module HOT 1
- About the box down-sample kernel HOT 2
- About Vimeo Dataset. HOT 1
- Not all septuplets listed in "sep_testlist.txt" and "sep_trainlist.txt". HOT 1
- 2x model
- Issue in SSIM implementation
- License of Vimeo-90K?
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from toflow.