Archive

Archive for September, 2012

NO QUEST Meeting this week, Sept 28th

September 26, 2012 Leave a comment

Due to travel schedules and workload, there will be no QUEST Meeting this week.

From Capt Amerika –
the material that is quest related that I’ve been focused on
this week include the attached article on language of the brain (basically
the neural properties that lead to consciousness), and reviewing the work on
joint activity (klein work) – I was reviewing that work for a discussion on
cognitive electronic warfare that I’ve been having ongoing and their need
for quest like solutions –

The language of the brain

Advertisements

21 Sept discussion between Kirk, Adam and Capt Amerika

September 22, 2012 Leave a comment

21 Sept discussion between Kirk, Adam and Capt Amerika:
a. We’ve recently been attempting to capture where we fit by adding a confabulation layer to complement the Endsley model’s ‘projection’ layer – although the projection layer as presented by Dr. Endsley does confabulation – one approach is to combine what she calls projection with other functional needs we have (recollection = imagine what might have happened in the past versus recalling sensory data and what we call ‘perception = imagined present’ – not to be confused with the Endsley ‘perception level one’, imagine what might be happening now). The point being we’re interested in a set of processes that confabulates a representation (simulation) that allows the agent to ‘imagine’ what might be going on now = ‘perception’, what might happen next = imagination, or what might have happened in the past = ‘memory’. This view when blended with the ‘below the level of consciousness’ processing (implicit/ intuitive level) we hope will allow us to build far more capable agents that could respond to the unexpected query / environmental changes (that current schema/mental model approaches fail at). One purpose of this idea is to retrain the implicit response system. We’ve internally been using the phrase ‘situational consciousness’ versus ‘situational awareness’ to capture the distinction – SA is all about mutual information between reality and the internal representation where SC is more about generating a representation more useful for exploitation (stable, consistent and useful) versus matching to prior learned responses.

b. We have prior discussed properties of consciousness:

i. Various properties of human consciousness can be identified at the neural, behavioral, and phenomenal levels.

1. Neural correlates of human consciousness include the presence of thalamocortical signaling, fast, irregular, low-amplitude electroencephalographic (EEG) signals, and widespread cortical activity correlated with conscious contents.

2. At the Behavioral level consciousness has been associated with behavioral flexibility, rational action, and certain forms of conditioning. These can be related to cognitive properties involving widespread access and associativity, multiple discriminations, and the capacity for accurate reporting. These properties can be mapped to a variety of functions related to consciousness.

3. At the Phenomenal level, human consciousness involves the presence of a sensorimotor scene, the existence of a first-person perspective, the experience of emotions, moods, and a sense of agency.

c. Our SC process includes those properties of Consciousness not included in the current Endsley model for SA, for example the phenomenal properties. Also as stated in (a) above it also is focused on generating not just the information for projection but also for recollection and the imagined present. It does these in the context of using a multi-modal sensorimotor scene, the first-person perspective, emotion / affect, moods and a sense of agency / free-will. It is our position that these additional properties are critical to achieving the engineering performance necessary for a class of problems current solutions fail at.

Weekly QUEST Discussion Topics and News, Sept 21

September 21, 2012 Leave a comment

QUEST Discussion Topics
September 21, 2012

1.) The first topic this week is a review of what we communicated to Dr. Mica Endsley, SA Technologies, during her visit this week with respect to attempting to position our interests in the context of her model / definition of situational awareness.
We’ve recently been attempting to capture where we fit by adding a confabulation layer beyond the Endsley model’s ‘projection’ layer – although the projection layer as presented by Dr. Endsley does confabulation – one approach is to combine what she calls projection with other functional needs we have (recollection = imagine what might have happened in the past versus recalling sensory data and what we call ‘perception’ – not to be confused with the Endsley ‘perception level one’, imagine what might be happening now). The point being we’re interested in a set of processes that confabulates a representation that allows the agent to ‘imagine’ what might be going on now = ‘perception’, what might happen next = imagination, or what might have happened in the past = ‘memory’. This view when blended with the ‘below the level of consciousness’ processing (implicit/ intuitive level) we hope will allow us to build far more capable agents that could respond to the unexpected query / environmental changes (that current schema/mental model approaches fail at). One purpose of this idea is to retrain the implicit response system. We’ve internally been using the phrase ‘situational consciousness’ versus ‘situational awareness’ to capture the distinction – SA is all about mutual information between reality and the internal representation where SC is more about generating a representation more useful for exploitation (stable, consistent and useful) versus matching to prior learned responses. One approach I would like to discuss with the group is a nested Endsley model – if you will a SA system (with the Endsley levels monitoring the SA system that is working with the sensory data – if you can imagine the ‘confabulation layer’ we’ve proposed as being implemented to do ‘perception’/’comprehension’/projection of the SA system).
2.) The second topic that has consumed our bandwidth this week is a review of our prior presentations on Gists. The material includes review of an article and ideas from Prof Bar, Harvard; work with our colleagues from The Ohio State University – Stephens, Woods, Morrison; some work on the neural circuitry of first impressions/gists.
3.) The third topic that we’ve accumulated since our last presentation is a revisit of the material associated with making automation a team player. We’ve recently had a series of discussion with colleagues attempting to achieve Cognitive Electronic Warfare and exposing them to the ideas in the Klein / Woods article on challenges for joint activity.
4.) The last topic that we’ve been pursuing is associated with a train of thought associated with the ideas we’ve pursued after our discussion on whether the internet could become conscious. Specifically we want to extend the discussion on the use of modern approaches to processing ‘big data’ (Hadoop and the Google approaches that came after like Caffeine and Pregel etc.). This topic has the potential to combine several of the above ideas where the indexing step could be a form of Gisting.

QUEST Discussion Topics Sept 21

Categories: Uncategorized

Weekly QUEST Discussion Topics and News, Sept 14th

September 13, 2012 Leave a comment

QUEST Discussion Topics
September 14th 2012

We have proposed engineering of an artificial sys2 (the conscious part of a representation) using modern approaches to generating virtual worlds (especially those that allow for multi-modality interactions. The idea is that even though there is nowhere in the meat where all the information is ‘projected’ for a homunculus to be exposed to the information (I’ll use the term ‘script’) necessary to generate the VW can be the vocabulary used for deliberation over the content of the VW. This is one approach to computing with ‘perceptions’ as a complement to computing with sensed data (which we will still need to do with sys1 processes and where most of our current pattern recognition and data driven AI work lies).

This week we will have one of our colleagues, Matt Mcclure present to us the means to generate realistic virtual worlds

Our interest is to demonstrate the generation of what is consumed by the human sensory system as a really detailed virtual environment with a modest amount of ‘programming’ (lots under the covers but because it has been abstracted up a lot of the detail stuff is now provided by canned routines) – the point being that since you can generate all this detail with relatively little high level ‘programming’ – making the point that a system that analyzes the ‘programming’ at that high level could deliberate as if it had the full-fledged simulation
There is a continuum of automation, tooling, and raw artist work that brings VWs to life. Every environment or simulation will fall somewhere different on that continuum. Actually demonstrating the CryEngine we hope will be of some value. Discussing AFRL’s approach in simulation, but also how the game industry creates these worlds for “fun” could be interesting.

Matt will be presenting a basic overview of the process for creation of a virtual environment. Discussion includes:

Design Process
Design Principles
Demonstration of CryENGINE 3 Editor
World Creation
Terrain Creation and Modification
Terrain Texturing
Vegetation
Entities and Assets
Scripting AI

This is not a presentation on coding, but rather game and immersive environment design.

QUEST Discussion Topics Sept 14th

Invited Speaker – Dr. Mica Endsley

September 13, 2012 Leave a comment

INVITED SPEAKER: Dr. Mica Endsley, President SA Technologies

DATE: 19 Sep 12

TIME: 0930-1030

LOCATION: Bldg 441 Auditorium with VTC to Ft Sam Houston

TOPIC: “Cognitive Measures and Models for Persistent Surveillance”

ABSTRACT: Although advances in Intelligence, Surveillance, and
Reconnaissance (ISR) technologies are enabling access to a greater
variety of data sources, the deciding factor is how human analysts will
be able to capitalize on this enormous volume of available information
to support their decision-making. The extreme volume of data and the
speed at which that data flows can rapidly exceed human cognitive limits
and capabilities, and errors or delays in processing the data to develop
an understanding of its significance can undermine our goals in this
domain. The ability to achieve fast and accurate situation awareness
(SA) in the face of this data overload is a key factor that allows for
effective decision making and information exploitation. This
presentation will focus on defining situation awareness as it relates to
the USAF Global Integrated ISR mission, and will present a model of
situation awareness including the factors that act to both limit SA and
that allow for people to achieve high levels of SA against these
constraints. Based on this model, methods for improving SA through
training and system design will be discussed, along with measures for
assessing SA that can be used to validate new systems or training
programs. These tools and methodologies provide a strong foundation for
improving situation awareness in USAF operations.

BIOGRAPH: Dr. Mica Endsley is President of SA Technologies, a cognitive
engineering firm specializing in the analysis, design, measurement and
training of situation awareness in advanced systems, including the next
generation of systems for aviation, air traffic control, medical, power
and military operations. Dr. Endsley received a Ph.D. in Industrial and
Systems Engineering from the University of Southern California. Prior to
forming SA Technologies she was a Visiting Associate Professor at MIT in
the Department of Aeronautics and Astronautics and Associate Professor
of Industrial Engineering at Texas Tech University. She is the
out-going President of the Human Factors and Ergonomics Society and has
served on the USAF Scientific Advisory Board. She has given similar
invited presentations at the National Reconnaissance Office and at the
NASIC. She has authored over 200 scientific articles and reports on
situation awareness, decision making and automation and is recognized
internationally for her work in the area of situation awareness. She is
co-author of Analysis and Measurement of Situation Awareness and
Designing for Situation Awareness.

Categories: Uncategorized

QUEST Discussion Topics and News, September 7th 2012

September 6, 2012 Leave a comment

QUEST Discussion Topics
Sept 7, 2012

We will start this week reviewing where we left off last week – the Cambridge Declaration of Consciousness. Among the statements we would like to discuss:
In humans, there is evidence to suggest that awareness is correlated with cortical activity, which does not exclude possible contributions by subcortical or early cortical processing, as in visual awareness. Evidence that human and nonhuman animal emotional feelings arise from homologous subcortical brain networks provide compelling evidence for evolutionarily shared primal affective qualia.
We declare the following: “The absence of a neocortex does not appear to preclude an organism from experiencing affective states. Convergent evidence indicates that non-human animals have the neuroanatomical, neurochemical, and neurophysiological substrates of conscious states along with the capacity to exhibit intentional behaviors. Consequently, the weight of evidence indicates that humans are not unique in possessing the neurological substrates that generate consciousness. Nonhuman animals, including all mammals and birds, and many other creatures, including octopuses, also possess these neurological substrates.”

Next we would like to return to the discussion on computational theory of perception – computing with qualia and pre-qualia. What do we want to accomplish by this form of computation? What sort of problems do we envision solving that currently are problematic? Kirk and I had a related discussion last week and I realized we had not finished this discussion.

The last topic we have on our agenda for this week is to continue our literature review on our CHLOE proposal axes. Specifically we want to return soon to the issues in making automation a team player and to do that we will use the formalism of the Klein article on that topic. We specifically want to ask the question what in our QUEST formalism addresses the challenges he lists.

QUEST Discussion Topics Sept 7