luca-m / emotime Goto Github PK
View Code? Open in Web Editor NEWRecognizing emotional states in faces
Recognizing emotional states in faces
Import images from http://www.mmifacedb.com for validate models trained with CK+.
Select emotion tagged videos, extract frames and put them in proper dataset folder.
Develop a better GUI can show image processing intermediate processing steps for a better understanding of illumination issues.
Hi,
Can you help me with this error?
./emotimegui_cli ../resources/haarcascade_frontalface_cbcl1.xml none 48 48 3 5 4 svm ../dataset_svm_354_cbcl1_1vsallext/classifiers/svm/*
OpenCV Error: Bad argument (The SVM should be trained first) in CvSVM::predict, file /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_graphics_opencv/opencv/work/opencv-2.4.11/modules/ml/src/svm.cpp, line 2131
libc++abi.dylib: terminating with uncaught exception of type cv::Exception: /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_graphics_opencv/opencv/work/opencv-2.4.11/modules/ml/src/svm.cpp:2131: error: (-5) The SVM should be trained first in function CvSVM::predict
Abort trap: 6
Thanks
I'm on windows, compiling with mingw and all installations seems to be good (assets folder is populated). If I try
python gui.py --cfg ..\dataset_svm_354_cbcl1_1vsallext\dataset.cfg --mode svm ..\dataset_svm_354_cbcl1_1vsallext\classifiers\svm*
It doesn't works and it prompt me: ERR: something wrong (No section: 'TRAINING')
What's wrong? Thanks
Hello
First of All thanks for your project. I am working on emotimevideo_cli on Ubuntu 14.04 opencv 2.4.9
And I am giving multicolor mp4 video for input but emotimevideo_cli says "Loading 'video.mp4'" and "init done""opengl support avaliable"and then return below error.
Opencv Error: Sizes of input arguments do not match (the sample size is different what has been used for training )in cvPreparePredictData
Aborted core dumped.
I cannot resolve the problem What do you offer me to do? Is it about video size, color or framerate?
Hi,
I'm trying to run your code on Windows and am running into trouble. I can compile and run both the webcam and video file versions of your code. However, when I do run the code, a blank window appears (no video is shown -- see attachment for picture of problem). I have tried several video files with faces clearly in them. I have also used my head against a blank backdrop with a webcam, and the problem persists.
This is the line that I am using in MVS 2012:
emotimegui_cli C:/opencv/build/emotime-experimental-v1.1/resources/haarcascade_frontalface_cbc1.xml none 48 48 3 5 4 svm C:/opencv/build/emotime-experimental-v1.1/dataset_svm_354_cbcl1_1vsallext/classifiers/svm/*
If running in MVS 2012 isn't optimal, how can I run this line in a command prompt on Windows?
I have tried your instructions here: https://github.com/luca-m/emotime/releases/tag/v1.1-experimental. However, they don't work for cygwin on Windows (possibly because ELF executable files don't run on Windows machines).
Deploy the project (zip + examples link + README + doxygen dox + report)
gui modules and tools which shows current frame and predictions
Hello!
I cloned your project and ported it to Android, so it now takes images from the camera and runs analysis on them to see expressions. However, I keep getting "Surprised" and Score=1.0. Could you tell me what size I need for the image (I am currently doing 48x48) and whether or not the image needs to be monochrome before I pass it to the FacePreProcessor?
Thanks,
Rohit
I'm so interested in this project, but I encountered some problems while trying to use it. After successfully compiling the project, I have trouble on how to open the GUI.
I tried running the project on Visual Studio 2012, but it won't appear.
Which part did I do wrong?
Thanks in advance :)
modules and tool for prediction with svm
Tool and modules for prediction using adaboost
Gentle Ada Boost regression trees (no neutral): a3d381c
My PC doesn't support the featselect_cli
utility due to lack of RAM (it require over 3GB of reserved RAM). Do someone with an ultracool, finetuned notebook run it on all the ada classifers trained?
Error when I try to use:
python datasetInit.py --cfg test_dataset.cfg dataset_folder
dataset_foldet = the empty folder
test_dataset.cfg = an exact copy of example_dataset.cfg
training and validation test are correctly populated with subfolder, classifiers folder show only svm.
It seems that it doesn't create ada subfolder and that all the process is stopped.
Thanks.
Adjust the readme for better integration with Doxygen
hi ,i am trying to run your project .for some time it was working fine but after a few trys it is giving some errors like this and i request you to help me out.
errors:
OpenCV Error: Parsing error (../dataset_svm_354_cbcl1_1vsallext/classifiers/svm/(0): Valid XML should start with '') in icvXMLParse, file /build/buildd/opencv-2.4.8+dfsg1/modules/core/src/persistence.cpp, line 2252
terminate called after throwing an instance of 'cv::Exception'
what(): /build/buildd/opencv-2.4.8+dfsg1/modules/core/src/persistence.cpp:2252: error: (-212) ../dataset_svm_354_cbcl1_1vsallext/classifiers/svm/(0): Valid XML should start with '' in function icvXMLParse
Aborted
Hey! First of all, thank you so much for making this project. Your sources and cited papers have taught me so much more than I expected to learn when I started my own.
I'm trying to build emotime on Windows, but I simply cannot figure out how to include the opencv_contrib directory in the make file. Could you help me out with that?
Inside INPUT, set the directory you are currently working on. Otherwise the warnings will be impossibile to read. Once the documentation is reasonable warning-free, commit the doc and the doxygen edit
Review CvMLData, make it works or load directly feature matrix, transform it in a feature vector and manually prepare training set (and validation set)
what is the meaning of this error
Tune up face detection parameters for better detection performance
Gabor parameters:
May work better
If I Ctrl+C out of the emotime_gui program, my camera stops working. I found this is because of this, so ill make a fix and do a pull request.
https://lawlorcode.wordpress.com/2014/04/08/opencv-fix-for-v4l-vidioc_s_crop-error/
Fill dataset: I'm not sure that I'm giving the right path, but it seems works... expect the error.
Thanks
c:\workspace\emotime\assets>python datasetFillCK.py --cfg test_dataset.cfg dataset_folder C:\workspace\emotime_dataset\ck\ck_plus\cohn-kanade-images C:\workspace\emotime_dataset\ck\ck_plus\Emotion
INFO: 123 subjects found in CK+ database
INFO: Processing subject S005
INFO: Processing shot 001
INFO: Picture 'S005_001_00000011.png' has been marked as disgust
ERR: something wrong ('module' object has no attribute 'next')
Some python scripts need a slight edit.
I ran a brief test yesterday but I still have problem with this issue.
[FOLDER]
...
# Folder containing the csv files of the features for training set and validation set (one file per classifier)
TRAIN: trainfiles
...
I miss the point. "trainfiles" is another folder where to put the csv. data?
Do we create this one with datasetInit.py? because right now "trainfiles" doesn't exist...
What does it change the name of the folder?
Last thing, in my previous test I failed one test with facecropping on neutral, is it possible that this result have some bad influence on the rest?
Thanks
Enanche face detection using CLAHE for local equalization.
New emotime version requires [TRAINING]
and [VALIDATION]
configuration sections. The dropbox configuration needs to be updated.
./emotimegui_cli ../resources/haarcascade_frontalface_cbcl1.xml ../resources/haarcascade_eye.xml 52 52 1 5 8 svm ../svm_1vsall_1_5_8_95c2eb0b58/*
this errors comes up. I have downloaded the trained models from https://github.com/luca-m/emotime/releases
when i m running this command ...
following messages shows up....
VIDIOC_QUERYMENU: Invalid argument
VIDIOC_QUERYMENU: Invalid argument
VIDIOC_QUERYMENU: Invalid argument
VIDIOC_QUERYMENU: Invalid argument
VIDIOC_QUERYMENU: Invalid argument
VIDIOC_QUERYMENU: Invalid argument
VIDIOC_QUERYMENU: Invalid argument
and sometime this work quiet slow...
and you had give the link of version 1.2 in that 1vsall and 1vsallext link are not working....
Change the gui files to the standard C++ conventions (in refractoring branch)
Train SVM
We have to
To describe classes and strategies (without details about parameters and results).
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.