From BrainStorming to BrainDriving

An impressive application of cognitive science and robotics from the AutoNOMOS Labs of the Artificial Intelligence Group of the Freie Universität Berlin.

BrainDriver

MadeInGermany controlled via the Emotiv EPOC Brain Machine Interface:

After testing iPhone, iPad and an eye-tracking device as possible user interfaces to maneuver our research car named “MadeInGermany”, we now also use Brain Power. The “BrainDriver” application is of course a demonstration and not roadworthy yet but on the long run human-machine interfaces like this could bare huge potential in combination with autonomous driving. For example when it comes to decide which way you want to take on a crossroad while the autonomous cab drives you home.

PANORAMA 3D shape descriptor applied in medical image analysis

To my pleasant surprise, a study on medical image analysis drew my attention today that found particularly useful my earlier work on 3D shape descriptors for 3D shape matching, and in particular the work described within the paper "PANORAMA: A 3D shape descriptor based on panoramic views for unsupervised 3D object retrieval".

The respective work titled as "3D Shape Analysis of the Knee Extensor and Flexor Muscles in Patients with COPD using Mesh Projection-based Features", by Hengameh Mirzaalian, Ghassan Hamarneh, Bahareh HajGhanbari and W. Darlene Reid, used and extended the PANORAMA descriptor in order to match patients MRI of thigh muscles against a database of muscle instances.

Referring to their paper:

"In order to differentiate 4 individual thigh muscles in the healthy versus COPD groups, we first applied a state-of-the-art 3D shape descriptor; the WT-based shape descriptor proposed by Papadakis et al. [14] resulting in cylindrical projections. A comparison between the classification accuracies obtained by the aforementioned descriptors and the global shape descriptors by Ward et al. [20] and HajGhanbari et al. [10] shows that, averaged over all the 4 muscles, the WT-based shape descriptors outperformed the global shape descriptors."

"Although the presented descriptors were applied to differentiate thigh muscles, they might have a widespread application for other conditions and chronic diseases that result in muscle atrophy such as chronic heart diseases, AIDS, cancer, and osteoarthritis."

Although the development of PANORAMA aimed toward effective retrieval of generic 3D objects, it makes me proud that it could further serve the domain of medical image analysis.

I feel that this is the essence of good research, that is, to promote and develop tools and methods that assist humans and science in general in a cross-disciplinary fashion.

Robots Manufacturers

I assembled a draft list of robots manufacturers, quite popular within the robotics community as a reference for what is currently the market demand together with the respective products.

Willow Garage

iRobot

Aldebaran

Mobile Robots

BlueBotics

KUKA
ROBAI
ENERGID
ST ROBOTICS
LYNX MOTION
MOTOMAN ROBOTICS
ABB
Crust Crawler
Kraft Telerobotics
Robotiq

In time, i will try to keep this above list up-to-date and i would certainly appreciate your feedback.

I would distinguish the following prominent categories of robots:

• Rescue Robots.
• Military Robots.
• Home Robots.
• Service Robots.
• Research Robots.
• Entertainment Robots.
• Space robots.

And i will finish this post, with perhaps the most famous robots song (and quite old).

Sebastian Thrun's talk on Google's driverless car

A demonstration of the potential of the applications of machine vision and robotics. The talk is given by Sebastian Thrun, pioneer in the field of robotics research.

The Methological differences between human and machine vision

From "3D Shape, Its Unique Place in Visual Perception", Zygmunt Pizlo

Resources for Vision and Robotics

For the practitioners of Vision and Robotics, relevant sources of information and software platforms are the sine qua non for effectiveness (doing the right thing) and efficiency (doing things right) in research and development.

A full list of resources would be of course boundless, but the prominent (in terms of popularity and spread) resources would probably suffice the average interest:

Mailing lists (conferences cfp, job openings, seminars, news):

http://lists.diku.dk/mailman/listinfo/imageworld (Computer Vision community)

https://lists.iais.fraunhofer.de/sympa/info/euron-dist (Robotics community)

http://www-clmc.usc.edu/mailman/listinfo/robotics-worldwide (Robotics community)

Software platforms:

Robot Operating System (ROS): http://www.ros.org/wiki/

Open Computer Vision Library (OpenCV): http://opencv.willowgarage.com/wiki/

Computational Geometry Algorithms Library (CGAL): http://www.cgal.org/

Blender (3D Content Creation Suite): http://www.blender.org/

Open Graphics Library (OpenGL): http://www.opengl.org/

Web Graphics Library (WebGL): http://www.khronos.org/webgl/

The use of all the above resources are FREE for everyone and tutorials are all around the internet for amateurs and professionals. Get a grip on and enjoy!

Welcome

This is my initiating post of my blog "Vision & Robotics".

I created this blog as a means to communicate news, advances in the science of machine vision and robotics as well as promote research in these fields.

The interest as well as imagination of people for automatic, self-intelligent machines traces back thousands of years ago. To the best of my knowledge, "Talos" (or equivalently "Τάλως" in greek) was probably the first humanoid robot constructed by God "Ifaistos" to guard the island of Crete according to the hellenic mythology.

The extent to which the records about Talos are true are not so important. What is important is that man's creations first take shape in imagination and in this aspect, we can accept that this first step was made a long ago and that today we have a plethora of tools in our disposal to make the trip from imagination into reality.

Welcome :)