For my COGS 310: Consciousness project, I’ve been researching synethesia. Synethesia is the experience of a sense cross-modally. A synesthete may perceive time in a spatial rather than numeric terms. This particular article I’ve been reading is about an individual, R, who perceives his personal emotional states with colour. I drew up some of the juicier details of the study:

Subject R

R is a colour-blind synaesthete. Visually, he has trouble discriminating between between certain shades of red, brown, and green. R’s sensation of colour is, seemingly, a quale.[^1] The paper discusses this aspect of R’s perception as being an aura or aura photism.

R’s Perception

When R encounters a person or object, he states that he experiences a sensation of colour: one that is never projected externally onto the person being perceived. Like all synaesthetes, R claims that his experience is highly consistent over time and cannot be suppressed by will.

Experiments & Method

The first experiments were establishing the extent of colour perception in R. They involved colour identification tasks and a stroop test. He had a hard time with certain colours and greys (especially red-green discrimination). There was no significant Stroop effect. The next set of experiments were aimed at inspecting the consistency of R’s aura photisms. The authors used the IAPS [^2] and went through approximately 500 images, naming an aura for each one. His responses were 98% consistent over time when he repeated the test 30 days later. The final set of pilot experiments involved the colours of emotion. Some of the stimuli that R described as perceiving a green photism included: tarantula spider, mutilated human head, mutilated dead body, half naked man with badly burnt body. It would seem that R’s green photisms were evoked with emotionally negative pictures.

I just stumbled on a great library for neural net architecture at PyBrain. I haven’t had a chance to really get my hands dirty, but just looking at the algorithms available, I really want to give it a go with a couple projects:

• Backprop
• Rprop
• Policy Gradients
• Support Vector Machines
• Evolution Strategies
• CMA-ES
• Competitive Coevolution
• Natural ES
• Natural Actor-Critic
• Reward-weighted regression
• Evolino
• Fitness Expectation Maximization
• SPLA
• PCA/pPCA
• LSH for Hamming and Euclidean Spaces

In finding new ways to describe and emulate cognitive microstructure, self-organizing maps (SOM) offer insight into how the visual system might work in a connectionist model of the brain.

For this example, I am going to use len’s examples and code to show how artificial neural networks can be applied to a visual emulation of the brain.

Note: This is a repost of a previous entry.

Of the many great experiences I had in Syria, some of my favourites were at family gatherings and parties. Of course, the sort of reunions that happen in the “Old Country” are far more lively than the usual Christmas get-togethers that North Americans participate in. For one thing, because I was the only one from my household in Utah to visit family in Syria, I was constantly taken from relative to relative, meeting people aged from less than a month to 80+ years. Family in Syria means not just one’s aunts and uncles, but also their children’s relatives. It means that I not only met people distantly related to me by 2nd or 3rd generational gaps, but also be met every friend, distant relative, and acquaintance of everyone else’s friends, relatives, and acquaintances. I probably met 100+ different people from a wide array of backgrounds and professions: from business to dentistry. The family dynamic in Syria is broad, to say the least.

With the meeting and greeting of so many people, the same questions always pop up, “How’s your family? How old are you now? Where are your parents living now?” ; etc.

However, there was one question that always befuddled me, not because I didn’t have an answer, but because the answer always took time to explain: “What do you do right now?”

This question I’d always answer with a similar response; something along the lines of, “I’m a student of Cognitive Science”.