lucam

the models are openCV models, so yes it could be possible, but _at least_ you need to re-implement the multi-classifier logic (1-vs-many or many-vs-many). PR are welcome

Could you try with opencv version 2.4? Il 02/apr/2017 17:09, "mazzalabs" ha scritto: > I'm getting the same problem. > Cloned again opencv and opencv_contrib by tag 3.2.0 for both...

Which version of opencv are you running? At the moment 2.4 have been tested. Sent form my phone. Il 13/feb/2017 19:33, "Sundar Raman G" ha scritto: > The follow error...

Any error? Could you post here the command you are running? Il 26/nov/2015 03:13, "Nik Bist" [email protected] ha scritto: > Upon running the python gui.py ..., I get the message...

../ in windows ?

We used adaboost in very early dev stage. At the moment my advice is to use svm, but we are also thinking to give a try with some random forest...

Well, if the windows open and no image is shown it means no face is detected. The gui is quite minimal and loads the image after a successful image processing...

However i can give a try. Are you using the master code and the models in https://github.com/luca-m/emotime/releases/tag/v1.2-experimental right?

``` 42: assert(!cascade_f.empty()); ``` means the face cascade classifier must be loaded correctly. Few lines abole il cascade_f is loaded from the specified files. maybe the file has something wrong...

However I've trained a some new models that works better with the current master branch. [Here](https://copy.com/c4rOYmPgAYrdxdoS) . I'll add the link in the readme soon.