Graphic Impulse

Meredith Burgess Summer 2016 UROP Research Project. Pole Dancing Meets Live, Interactive Technology. To be debuted at Pole Theatre USA August 19, 2016

[graphicimpulse-blog]

A few stills from practice the other night! 

[graphicimpulse-blog]

View this post on Instagram

It’s funny how the process goes...the timing here just happened to be right so that when I came off the pole, the colors changed. It was a coincidence, but I liked the effect so much I ended up coding it in! 

[graphicimpulse-blog]

Practice timez. 

From last Sunday. 

[graphicimpulse-blog]

CU student Meredith Burgess brings tech to pole dance

Meredith Burgess is where pole dancing collides with computer science. You might not put the two together, but for the Boulder woman, the combination of her two seemingly disparate passions has brought her to the largest pole dancing competition in the world.

So my project ended up getting coverage in the local newspaper! Sweet! 

The link to this blog was mentioned at the end which is funny because I never brought that up and have sadly not been staying up to date here as much as possible. 

Sooo gonna try to do that. Yea.

[graphicimpulse-blog]

Back at it this week. Here is some particle emitter and collision detection work

[graphicimpulse-blog]

Been working as much as possible this past week, amidst circus and work madness. 

Reworking detection to work on a gradient rather than just discrete values, as reflected by the color changes. This was hardest for the hand gestures, as there really is nothing built for in between gestures. While doing this I discovered my velocity calculations were all kinds of incorrect, so hopefully it’s more accurate now. Also trying to take account for relativity due to distance from the camera. There’s been a lot of hacky math going on in my notebook. 

I also have started incorporating a particle system! Hooray!

Also, timing. It’s happening! The appearance of the squares is set to the musical tones at the beginning of the song. I’m going for a little bit less interactive with a few things just to make sure they happen no matter what. Gotta leave room for some error up there, I only get one shot. :/

A little stressed because I still haven’t done a full run through of anything yet. It’s very unlike my process to leave things up in the air. I like it locked down by now. I just keep telling myself I’m good at improvising and spontaneity is half the point of this piece sooo.... Agh! This is an experience all right. 

[graphicimpulse-blog]

Some neoplasticism madness. 

[graphicimpulse-blog]

Whoa! 

I actually made all of my goals! It’s Tuesday and I already have all the detection systems I said I would by now. They could be more refined yes, but I’m still proud of myself. I have not looked into mapping yet. Will do for the rest of the week. 

But! All this work gave me an idea. Give me a day and we’ll see if it works out. 

[graphicimpulse-blog]

 I restarted some things, got a basic layout of an overall motion sensing/detection demo going. I got caught up in trying to figure out data binding for text for awhile. It seems straight forward enough with the sensor status included in the sample code, but once I got to the hands I became very confused. I decided since I don’t actually care about text showing up on screen, just feedback on hand state, I’d keep it simple with colored rectangles. If I wanted to I could probably make images that showed up with text in them. Somehow I think that’d be easier than data binding a string to a text box.  

I swear to god, some things are so hard for no reason. 

Tomorrow I’ll try to add in velocity again. 

[graphicimpulse-blog]

Redirecting

Particle systems are making my head hurt. Especially when the only apparent WPF particle system engine that exists only exists in a video demo online. Can’t find the source anywhere. Gah.

Going to try to focus on the sensing for the next 5 days and worry less about the output. Here are my goals for next Wednesday, so I can have ~something~ of a demo for my labmates:

Map body index silhouettes to real size human during projection (probably going to be the most challenging thing)

Detect body in XYZ regions (a 9x9 cube of space)

Detect left-right movements

Detect up-down movements

“Strikes” and “sweep” movements of hands and feet

Hand gesture recognition (already built in to the API, so this one is easy)

Acceleration/Velocity (already done)

Body rotation/orientation (profile, front/back, upside down?) 

Basically the goal is to build a gesture detection vocabulary so that I can combine them and build a more complex dictionary of movements. Ultimately I can decide different outputs for different combos and thus easily build the look of the entire piece in relation to the movement. Obviously the “what” of the output will come later...

[graphicimpulse-blog]

Feeling Stressed

The more I delve into this project, the less I feel like I understand. WPF seems to be a much more complex world than I realized. 

Trying to look into particle systems coding. I think it may be the answer to my needs. If only I could make heads or tails of it. 

[graphicimpulse-blog]

It took awhile to get this idea figured out, but hopefully it’s finally going somewhere.

I didn’t code at all last week because I was training 5 hours a day and spending the rest trying to recover. I intend to double down this week and really have a solid demo by mid-week next week. 

[graphicimpulse-blog]

Exploring animation capabilities of WPF forms, all written in C#. 

Woo, man, it took awhile to figure out how to animate multiple objects along a path calling the same function every time. Turns out, it was just one line of code that needed to be moved (the one to do with name scope, in this case). Every. Time. 

Ok, but there is some cooler stuff coming out of this. 

[graphicimpulse-blog]

I have the beginnings of something! How exciting. 

I added joint tracking back in so I could hone in on hands specifically. I’ve got little balls of energy that follow them around if closed or in the “pointing” position. Unfortunately my finger is currently injured, so my ability to make this gestures is unreliable. 

I’ve also got a somewhat hacky function to clear the screen so I can draw the outlines over each other and refresh it every time I become still. It’s also buggy, but I look forward to projecting this over me to see what it looks like. 

Right now to make both effects work I’ve got the bitmap overlaying the canvas. Unfortunately the two don’t align perfectly. That’s gonna take some work to get fixed. Bleh. 

[graphicimpulse-blog]

I finally got the outline thing figured out. I add all the determined outline points to a list of points, then in the frameArrived function, I determine 40 random numbers, add them to an array, and use those numbers as the indices of the body points I select to draw ellipse arrays around. 

Whew. So complicated. And it still only looks vaguely how I want it to look. 

[graphicimpulse-blog]

I spent a lot of time working with this today...I want it to conform to the outline I created yesterday, but that is proving to be a little trick right now. Nonetheless, it is an interesting effect and I think the more I work with it the sleeker I can get it to look. 

[graphicimpulse-blog]

Back from vacation this week. 

Feeling like I’m starting to understand this stuff a bit better. Set up some code that tracks left, right, upper, and lower extreme points, plus the average center point. Also figured out how to take the silhouette data and make it just an outline. I think this will open up some avenues. I don’t have to rely on the sketchy and sensitive kinect algorithms for finding specific body points, but can still use the outline and basic shape of my figure, which will be good for the convoluted movement I intend to have.