ci2cv / face-analysis-sdk Goto Github PK

When I use this project, I find that when the situation that one eye open and another close, the animation shows that both eyes close half.
How can I fix this bug? I have checked the code, and I rewrite some, but useless.
Can any guys help me?

I figured this out but wanted to put it here none the less with solution in case others run into it.
Was getting the following errors when building on (OSX 10.11.6):

/Users/username/Development/FaceAnalysisSDK/face-analysis-sdk-stable/src/map-list/main.cpp:71:21: error: 
      use of undeclared identifier 'fork'
    pid_t new_pid = fork();

/Users/username/Development/FaceAnalysisSDK/face-analysis-sdk-stable/src/map-list/main.cpp:86:16: error: 
      use of undeclared identifier 'execvp'
      int rv = execvp(command.c_str(), argv);

by adding #include <unistd.h> to src/map-list/main.cpp and then re-compiling I was able to build successfully.

Hello, everyone, can anyone find the paper?
M. Cox, J. Nuevo, J. Saragih and S. Lucey, "CSIRO Face Analysis SDK",
AFGR 2013.

Dear All,

is it possible to train a new model with my own data set?

i just what train the face tracking model which describe in the paper "Deformable Model Fitting by Regularized Landmark Mean-Shift", IJCV 2011.", Could you give me some tips about it.

thank you very much~ ^_^

There are three compiling and linking errors:

• osx-configuration.mm: missing NSValue.hpp
• worker-thread.cpp: character case error
• CMakeLists.txt: missing setting for GnuStep on linux

When will the windows build support release? I see there is an plan on http://face.ci2cv.net/doc/ .

Hello,

I have seen that there has been no commit after Jan 2014. Is this SDK still in active development?

I'm looking over the code, and I see that there are three branches containing fixes for build breaks, and two adding new features. None of the authors of these branches have submitted pull requests (yet), so it makes sense that they haven't been merged. However ...

Given that all of these are at least a month old, and some almost a year old ... isn't it time to spur activity? Seems like the fixes should be picked up. Specifically, I see:

https://github.com/tetard/face-analysis-sdk/ --fixes for apple & LLVM compiler
https://github.com/hansonrobotics/face-analysis-sdk/ -- fixes for compiling on debian wheezy
https://github.com/blackibiza/face-analysis-sdk/ -- fixes for Qt5

Next: the new features:
https://github.com/supermartian/face-analysis-sdk/ -- stuff for recognizing emotions.
Hard to tell how complete/incomplete/buggy this is.

https://github.com/iCogLabs/face-analysis-sdk/ -- beats me. Some kind of hacky stuff .. ??

./tracker/PRA.cpp
../tracker/FacePredictor.cpp \

Hi Gautam,

I tried your Face Analysis SDK.It's really fine and easy to use.
I tried the landmarks extraction for the image using face-fit (executables),then started tested all the executables.
I struggled on creating avatar for my input image.Pleas clarify me to make use of it Gautam.

My environment:

1. Ubuntu 14.0 LTS.
2. g++ Compiler.

Command to execute the Avatar executables(Expression Transfer):

./expression-transfer 3.jpg 3.pts tt107.jpg tt107.pts out.jpg

I am getting one mask image as output in the name of out.jpg, but the target face is not substituted to my source image. The Command i executed is right or not.Help me to clear my doubt.

Thanks in advance!Hope expecting the help from you

Regards,
Gopi Krishnan.R

Hi

How can I install face-analysis-sdk on Ubuntu and use it ?