About Me

I am Gregoire Milliez, a French PhD student, working on human robot interaction.

In 2006 I started my studies in preparatory school (La Martiniere Monplaisir, Lyon).

In 2008 I pass the "concours pour grandes ecoles" and started to study computer science and applied mathematic in ENSEEIHT. After my 2cd year, I went in Japan in 2010 for 1 year internship in CNRS-AIST JRL. I worked on HRP-2 platform and mainly on vision server and object detection.

In September 2011, I came back to France to finish my master degree in ENSEEIHT. I took courses from the multimedia master. In 2012, as my last year internship, I did 6 months in Honda Research Institute in Wako (Japan). I worked on a prototype to build a humanoid math teacher. To build this multimodal multiparty architecture I used several modules for image processing from kinect data, speech processing from multi channel microphone and situation recognition using Markov Logic Network.

After getting my master degree in computer science engineering and my master degree in multimedia in September 2012 and february 2013, I started a PhD in LAAS-CNRS. My subject is "situated dialogue for human robot interaction". My goal is to provide relevant data about the environment to feed the dialogue manager to produce socialy acceptable robot dialogue. I'm particulary interested in agents belief management to enhance perspective taking ability to the robot.



2010-2011 – 1 year internship at CNRS-AIST JRL in Tsukuba (Japan).

During this one year internship I had the opportunity to work on HRP-2 platform, mainly on developping a vision server framework and an object recognition plugin.

2012 – 6 months internship at HRI (Honda Research Institute) in Wako (Japan).

For 6 months I worked on a project to create an English teacher from a robot. The robot had to be able to interact and speak with several human. I mainly worked on image and sound processing using a kinect and multi-channel microphone. I also use Markov Logic Network to assess the situation according to the input data.

2013-2016 – 3 years PhD at LAAS-CNRS in Toulouse (France).

I am curently in my 2 year of my PhD. My goal is to make the robot socialy acceptable by the human by making it understand the situation and react through dialogue on a socialy acceptable manner. To do so, The robot uses the ability of perspective taking to infer the other agents belief state and react accordingly.

Center of interest

Robotics (of course ^^).

I really believe that robotic and specialy those involved in human robot interaction are the future. I really enjoy working on this field
I think one of the first way to evaluate an agent intelligence is through dialogue.


I like to take pictures. I am still a beginner but I really enjoy catching a moment. Specially portrait, catching an expression on a face. You may want to have a look at my tumblr.


I also really enjoy traveling, discovering new places, new cultures and new working methods. Generaly speaking, I really enjoy new experiences and I hate doing the same thing again and again.