Want to avoid being auto-tagged by Facebook, Google photos, flickr, and the like? Want to create a party environment for all your cohorts that ensures all attendees remain unrecognizable to the collective scrutiny of the bots? Whether you’re familiar with the Many-faced God or not, you can benefit from the dark magic that streams from its collective. Here’s how.

Safety Amongst the Heard

The approach of this projects is derived from a classic hacktivist tool – the DDoS attack. Dos stands for distributed denial of service. Essentially this tactic can shut down any service the way the Star Wars fans shut down Fandango when tickets went on pre-sale; by overloading the the servers dedicated to a service with requests, the service becomes effectively unavailable.

Typically this tactic is used to shut down web services of corporations that have been misbehaving or underestimating the power of the internet, but in this case we’ll be applying the concept to render Facebook’s auto-tagging feature effectively useless.

Making A Mask of Masks

To kick things off let’s just start with a ton of faces. After running this through Facebook’s tagging system I was surprised as how good a job it did.

Even though most of the faces were clipped or obscured, the tagging system was able to identify 18 / 22 faces. The anti-establihsment won’t settle for an 18% success rate.

To up the unsuccess rate we’ll add in some extra facial orifices.  Maybe we only need to mess with each face a little bit to throw off the recognitions algorithms.

The tagging system is still recognizing 17/22 faces, and 17/27 if you begin to count based on all the eyes, noses, and mouths present. This increases our un-recognition rate to between 23% and 37% for an average of 30%. Better, but nothing you want to trust with your social life. Let’s take things a step further.

For the next mask we’ll use the previous mask, but overlay a rotated copy of original image. The result is a nearly unrecognizable hot mess of facial features. Everyone becomes one, and one becomes no one. This pleases the Many Faced God, but what does the Facebook’s recognition think of our abomination?

One face detected, and he doesn’t seem to be too happy about it.  Still, this mask’s 54 faces brings it’s un-recognition rate to 98%. Not perfect, but certainly ready to begin expanding the applications at a responsible rate.

Human Trials

With no funding partners to sponsor testing subjects less conscious of their social media presence, I will have to sacrifice myself to the Many Faced God.

Just Walk Right In

The strength of this strategy over other facial recognition obfuscation techniques is that it does not require individuals to do anything irregular. No face paint, no fancy clothes or accessories. Instead, the #antitag environment protects the identity of anyone inside it. To achieve this level of obscurity, the masks created earlier are projected all over a room, so that anyone within the room gets a face full of lasers and can’t see for hours becomes covered in the obfuscatory facial features.

Mask of Masks 1.0

With red boxes indicating the recognition of an incorrect face, and a green box indicating the recognition of the correct face, we can see that this first mask has a 78% success rate at preventing recognition. We can do better.

Mask of Masks 1.1

Let’s skip to the last mask, just to get a sense of the results we can expect to achieve. This mask has a 100% success rate at preventing correct facial identification, and an 83%  success rate at preventing incorrect facial detection. 

Mask of Masks 1.2

Adding a third layer of faces achieves total obfuscation. Coincidently total obfuscation is the new Clinton campaign slogan. No faces recognized. With 82 faces present in one frame, effectively there are none. 

For my Pepper’s Ghost project I decided to piggyback off of an idea that Professor Novy sent out to the the class. The project was called the aspire mirror; when the user peered in to the mirror, it reflects what the user is inspired by. Here’s a link to the website.

http://www.aspiremirror.com

With a half-silvered mirror and a massive 8k tv, I had all the technology I needed to start creating my project. After studying a great deal about pepper’s ghost and seeing the aspire mirror, I realized I could use this method to create a new form of online shopping technology!

This project works best in a room with natural lighting. Unfortunately the 8k tv I had available was in dark room so I had to provide some lighting with artificial lights, but when creating this effect in my room with sunlight over my computer screen it worked much better. I first placed the half-silvered mirror over the 8k tv as closely to the screen as possible. On the 8k tv behind the mirror I created a black background. Since the background behind the mirror on the tv was darker than the light in front of it,the mirror reflected my image back at me.  I then found items for sale online like a shirt or sunglasses and cut their images out, placing them on back backgrounds. After some very very careful positioning. I flipped through a slideshow of images to present what the future of online shopping would be like! First I showed the main websites regular online shopping page. I displayed the items for sale through the mirror (This showed through the half silvered mirror like  a regular screen). Then I chose an item to model by showing the item placed on a black background; wherever the black background on the tv was, I saw a reflection but wherever I had placed the item, that showed through!

Some complications of this project were trying to get the half-silvered mirror to stay on the 8ktv (I ended up duck taping it up), placing the objects to fit exactly on the right spot on my body in the reflection, and getting the lighting to work in the 8ktv room. In the future I think a good addition to this project would be to add an Xbox kinect or some other device to track exactly where body parts are to overlay images on top of them.

Below is an image of the final output on the 8k tv. Better images hopefully to come soon!

One of the most alluring pieces of equipment I have ever encountered is the phoropter, which is used by optometrists during eye exams to assess a subject’s vision and ascertain that subject’s eyeglass prescription.  The following link offers a description of phoropters, accompanied with pictures: https://en.wikipedia.org/wiki/Phoropter.  As a young child, I recall being pleasantly intrigued by watching my optometrist whirl the lenses and prisms in the device around with ease to alter my view; the swift changes seemed somewhat magical to me.  During eye exams, optometrists place a phoropter over a subject’s eyes and ask the subject to read a series of letters both close up and far away.  They iteratively change the lenses and other optics in the device to deduce the optimal prescription of the subject.  However, to my knowledge, optometrists have not yet implemented the Pepper’s Ghost illusion in conjunction with phoropters.  Since the ghost image is fainter than the object itself, a Pepper’s Ghost eye exam would be more challenging than the traditional one.  This would prove especially valuable for accessing candidates to jobs with strict vision requirements, such as astronauts.

I made a basic phoropter using two circular cut-out pieces of white cardboard and four pairs of lenses.  The lenses are displayed and labeled in the picture below.  As distinguishable from the labels, the types of lenses used were thick positive spherical, spherical meniscus, meniscus, and thin positive spherical.  An interesting aside fact I learned after reading about lenses is that plus lenses are synonymous with convex lenses, whereas minus lenses are synonymous with concave lenses.  Plus lenses are prescribed to fix farsightedness and minus lenses are prescribed to fix nearsightedness.  The two thick positive spherical lenses shown on the far left side of the image refract incident light appreciably, and thus when looking through them, the view appears quite blurry.  I was able to see clearly through the other three pairs, so the thick positive spherical lens pair served as an outlier case.

I traced out the shape of each of the lenses onto the two cardboard wheels with a pencil, such that the center of each lens was 4 centimeters from the rim of the wheel.  Shown below is a picture of my tracings on the wheels.  Scissors were used to pierce through the center of these marked regions and cut the shapes out.  Additionally, I cut out the shape of the largest of my lenses, the spherical meniscus, in two locations on a cardboard box for viewing windows of the ghost image.

Prior to fixing the lenses in place, I cleaned them with rubbing alcohol, rinsed them off with water, and then dried them with wipers.  A hot glue gun was used to securely mount the lenses on the wheels.  A picture of the two wheels with lenses attached is given below.

With the phoropter constructed, I next needed to make the Pepper’s Ghost illusion and position the wheels around the viewing windows I established.  I used a cardboard box to house the object and piece of acrylic needed for the illusion.  Two small holes were cut through the box in the locations where the center of each wheel was to be placed; small holes were also cut through the center of each wheel, as seen in the above picture of the wheels with lenses glued on.  Two large screws were pushed through the small holes on the box and the wheels.  Below is a picture of the phoropter mounted to the box.

Duct tape held the acrylic piece steady inside the box.  The object selected to produce the ghost image was a desk clock.  A clock was chosen because it provided the subject a means of discerning information.  The visual challenge was to look through the phoropter at the ghost image and tell the time, based on the locations of the second, minute, and hour hands of the clock.  The overhead view of the components inside the box is seen below.

In order to clearly see Pepper’s Ghost with my system, the lights were turned off and an LED light bulb was placed overhead the piece of acrylic as an illumination source.  Two shots through two different lens pairs are illustrated below.  The first is through the thick positive spherical lenses and appears blurry and unclear; the view through the right lens is the actual clock.  The second is through the spherical meniscus lenses, and the ghost image of the clock is discernible through the left lens.  The time on the actual clock is 11:01 with the red second hand on second 42; of course, since the ghost image is a reflection, it is inverted to look like the time is 12:59 with the second hand on second 18.

You can find a tutorial for the 0$ augmented reality device I built here

By recording video in the overhead camera position, also referred to as the bird’s-eye view, I achieved a visual illusion featuring two actors who seem to defy gravity and temporarily float in air.  The video is shown below.

 

The video features two wizards dueling with magical wands.  Throughout the action sequence, both wizards perform a levitation spell that portrays his opponent to be suspended in mid-air with his feet removed from the apparent ground.  In actuality, the actors are laying on the floor with their feet against a wall (the apparent ground), allowing each to simply remove his feet from the wall when his opponent launches a successful attack.  A camera operator is standing on top of a tall ladder, approximately six feet from the ground to film the overhead view.  The height from which the scene was filmed was a crucial aspect, as beginning film takes demonstrated that shorter ladders, such as those only two or three feet tall, did not enable a large enough space to be visible with the camera.

Avidemux was used for video editing and Kdenlive was just used to add audio to the edited video.  I carried out editing with two separate programs because for some odd reason Avidemux did not allow me to insert audio; although, Avidemux has the capability of running audio and video together.  During filming, there was a point in which the wizard on the right-hand side pushed his feet off against the wall in a jumping motion.  He then rose from the ground and walked over to the left side of his opponent, laying down again in an upside down fashion with respect to the camera’s view, before casting a final levitation spell.  Editing with Avidemux was done to facilitate the appearance of a surprise attack.  After the jumping motion, I removed video content up to the point in which the wizard on the right-hand side was completely off the screen.  I then kept eight frames of the wizard off screen; when watched right after the jumping motion, it appears that the wizard has vanished.  I then cut out all video content from that point to the point in which the wizard has established his new position behind his foe, so that it appears that the wizard vanishes and instantly reappears ready to attack.  With Kdenlive I added music played with a harp to complement the gentle nature of levitation.

Having never directed a video before, I became aware that it is quite difficult to verbally communicate an action sequence in a way that everyone involved can fully understand.  Although I had written out a script, it was suggested to me that I make a series of drawings to distinctly highlight each individual step in the sequence.  Implementing this new approach proved successful and enabled the actors to practice the choreography in a less hesitant manner.

For this week’s project I did a composite of live film and an animated character. This character is a chicken that I had previously created in Autodesk Maya. To create the chicken I started with a full-scale character design picture, which I uploaded into Maya to use as reference while I modeled it with polygons. After modeling the character by taking basic shapes like spheres and stitching them together, I rigged the character (implemented a skeleton with joints) and shaded it.

To composite the video I first filmed myself running around, imagining the chicken chasing me. To stabilize this video I learned that Youtube actually has a stabilizing button which performs the stabilization for free just by uploading and editing through youtube. After stabilizing, I downloaded the video and exported this .mov file into a sequence of Tiff images (using Premiere) to import into Maya. After this I created a new camera object that pointed at an image plane that played the sequential tiff images so I could overlay my character on top. Finally I animated my chicken, lit the scene and rendered; adding sound back into this new composited video.

Some bumps I ran into included animating the chicken to appear as if it came out of the computer. Originally I wanted to animate the chicken out of the computer, but this proved more difficult than I imagined, so I used a particle effect explosion. I originally wanted this puff of smoke to be white, however I found in order for it to shine white I needed a light to shine on it, which would in turn shine on the chicken, making it way too bright. So I decided to make the smoke black to avoid blowing out the chicken.

Some adjustments I would like to make for the future include maybe some textures for my chicken and maybe some better lighting for more accurate shadows. I would also like to turn the smoke white, and try and animate the chicken as if it were coming out of the computer screen.

Video Recipe:

Ingredients:

• Camera
• Tripod (very important!)
• cutting board, knife, fruits/vegetables
• Printed image of a strawberry
• 2 helpful friends
• Adobe Primier Pro

How it’s done:

• Set the camera on the tripod and make sure it’s not moving.
• One of the helpful friends is responsible for clicking the record/pause button on the camera.
• Start recording, when a magic need to happen, tell the friend to pause the camera, while the other helpful friend is responsible for replacing the objects in the frame.
• It is very important that someone else will swap the changing objects, so you move your hands as little as possible between each shot.
• After all the recordings are done, choose you favourite editing software to cut, connect and insert audio.
• I used Adobe Premier, and there are many tutorials for that online.
• I also speeded most video parts to play at X2 speed, and added music I found on the free music archive.
• To hide the video cuts, which were very noticeable, I added the “ding” sound on each switch.
• I changed the strawberries color using the “paint bucket” effect in Premiere, described here

Bon appetit!