MACS logo Image of robot KURT3D
home about partners events MACS in the press publications downloads links private
           

Sitemap
Contact
Imprint

            Fraunhofer IAIS

Joanneum Research
			Graz

LiU-IDA

METU Kovan Ankara

OEFAI

Universität Osnabrück

  Contents:
  
Videos
  Slides
  Public Deliverables

Videos

Affordance cueing for a non-liftable test object, involving computational perception units.
Affordance cueing for non-liftable test objects
video format: avi; © Joanneum Research

Affordance cueing for a liftable test object, involving computational perception units.
Affordance cueing for liftable test objects
video format: mp4; © Joanneum Research

Reinforcement learning for affordance perception.
Reinforcement 
			    learning for affordance perception
video format: avi; © Joanneum Research

Reinforcement learning for affordance perception.
Reinforcement 
			    learning for affordance perception
video format: mp4; © Joanneum Research

VOCUS-RT attention system and mean shift filtering for tracking a non-occluded blue test object simultaneously in L and R camera image.
VOCUS-RT attention system and mean shift 
			    filtering for tracking a non-occluded blue test object
video format: avi; © FhG/Fraunhofer IAIS

VOCUS-RT attention system and mean shift filtering for tracking a yellow test object simultaneously in L and R camera image, where the object is partially occluded.
VOCUS-RT attention system and mean shift 
			    filtering for tracking a partially occluded yellow test object
video format: avi; © FhG/Fraunhofer IAIS

KURT3D in simulator MACSim, training for liftability affordance.
Video of robot KURT3D 
			    in simulator MACSim, training for 
			    liftability affordance
video format: mpeg; © METU-KOVAN

KURT3D in simulator MACSim, scanning test objects.
Video of robot KURT3Din MACSim, scanning test 
			    objects
video format: mpeg; © METU-KOVAN

KURT3D in simulator MACSim, acting upon traversability affordance. Here, traversability includes pushing pushable objects out of the way.
Video of robot KURT3Din MACSim, acting upon 
			    traversability affordance
video format: mpeg; © METU-KOVAN

KURT3D in real world, acting upon traversability affordance. It applies the same perceptual routines that have been applied in the previous video in MACSim, and that have been trained in the simulator only.
Video of robot KURT3D in real world, acting upon 
			    traversability affordance
video format: mov (Quicktime); © METU-KOVAN

KURT3D in simulator MACSim, grounding the lift operator that is employed in goal-oriented behavior.
Video of robot KURT3D 
			    in MACSim, grounding lift operator
video format: xvid; © University of Osnabrueck

KURT3D in simulator MACSim, performing action planning and plan execution.
Video of robot KURT3D 
			    in MACSim, action planning
video format: xvid; © University of Osnabrueck

KURT3D in simulator MACSim, performing action planning and plan execution with an erroneous world model.
Video of robot KURT3D 
			    in MACSim, action planning with erroneous world 
			    model
video format: xvid; © University of Osnabrueck

KURT3D in real world (demonstrator scenario), executing the same plan as shown in the simulator video 12.
Video of robot KURT3D 
			    in demo scenario
video format: divx; © FhG/Fraunhofer IAIS

Back to top

Slides from MACS Final Review

Overview of MACS project achievements (FhG/IAIS)
State of the art in affordance-related research (METU)
Perception of Affordances (Joanneum Research)
Visual Attention for Exploration (FhG/IAIS)
Learning of Affordances (OFAI)
Reinforcement Learning for Affordance Perception (Joanneum Research)
Alternative Learning and Planning Approach (METU)
Planning with Affordances (University of Osnabrück)

Back to top

Public deliverables (PDF)

D0.1.7Publishable Final Activity Report
D1.1.1Specification of SW development environment
D1.1.2Specification of module interfaces
D1.1.3Implementation of SW development environment
D1.2.1Simulation of KURT2 Platform
D1.3.1Integrated implementation of reference control system
D2.1.1Identification of architectural requirements of an affordance-based control
D2.2.1Evaluation of existing control architectures for using affordances
D2.2.2Development of an affordance-based control architecture
D2.3.1Implementation of the affordance-based control architecture
D2.3.2A specification for a propositional planner and its interface to the MACS Execution Control Module
D3.1.1Top-down and bottom-up symbol grounding
D3.1.2Affordance recognition from visual cues
D3.1.3Saliency detection with visual attention
D3.1.4Prototypical affordance based object detection for MACS scenario
D3.2.1Multi-sensor affordance recognition
D3.3.1Prototypical sensormotor based affordance recognition
D3.3.2Sensorimotor decision making and affordance recognition
D4.1.1A conference or journal article summarizing the results of task 4.1. (Survey of affordance-related research)
D4.2.1+D4.3.1Combined deliverable:
Tentative Proposal for a Formal Theory of Affordances
Tentative Proposal for an Affordance Support Architecture
Prototype: Affordance-Based Motion Planner

D4.3.2A software prototype for affordance support
D4.3.3A software prototype of the propositional MACS planning module
D4.4.1A software prototype for an affordance monitoring module with empirical testing using various MACS robotics platforms
D4.4.3An evaluation of the MACS planning module in the context of the MACS architecture
D4.4.4Submission of a conference or journal article describing the results of D4.3.4 and D4.4.3
D5.1.1Overview of existing affordance learning approaches
D5.2.1Implementation of unsupervised and reinforcement learning algorithms
D5.3.1Robotic learning architecture that can be taught by manually putting the robot through action sequences
D5.3.2Robotic learning architecture capable of autonomously segment action sequences into affordances
D5.3.3Robot prototype learning affordances through self-experience V2
D5.4.2Prototypical software for representing and learning visual affordance support
D5.4.5Outlook towards affordance usage observation and imitation
D6.1.1Specification of final demonstrator
D6.2.1Simulation model of final demonstrator scenario
D6.3.1Physical robot demonstrator and scenario
D6.4.1Report on experiment design
D6.4.2Report on experimental results in simulator
D6.4.3Report on experimental results in demonstrator
D6.5.2Industry Day

Back to top