Skip to content

Latest commit

 

History

History
652 lines (442 loc) · 23.8 KB

unsorted.md

File metadata and controls

652 lines (442 loc) · 23.8 KB
Raw

Todo

https://www.artforum.com/video/hito-steyerl-how-not-to-be-seen-a-fucking-didactic-educational-mov-file-2013-51651

https://www.middletonspectral.com/company/about-us/

Photos:

Case Studies

  • Sarah Parcak & PCA
  • Swiss Cheese

Unsorted Links (In Process)

This page borrows significantly from the Computational Portraiture Workshop curriculum developed by James George & Alexander Porter of Specular.cc.

GL put these back, 9/12/2015 (they're for mocap & bodies):

Camera on fishing trawler moving through blood (hard to watch), by the Harvard Sensory Ethnography Lab

Portrait from motion capture: Merce Cunningham & Open-Ended Group

Once you have a 3D scan https://twitter.com/albertomoss/status/634870714856660992 https://www.youtube.com/watch?v=3yyLIG935zc

Human Flesh Search

Robert Hooke, Micrographia

Camera shutter speed synchronized with helicopter blade frequency

Candice Breitz

Still Life Lost Childhood Object: https://www.youtube.com/watch?v=QgHsYOybXa0 Embarrassing Object: https://www.youtube.com/watch?v=7mxM6mNSv5s Desktop Monument: Never Seen, Never Will

Landscape https://www.youtube.com/watch?v=ktcd2ZMWqAU

Portrait https://www.youtube.com/watch?v=0posLr4TTiI

MoCap Movement Telephone https://www.youtube.com/watch?v=AneJi1wGYvw

https://vimeo.com/134973504

RGBD

More

Uncanny Valley

RGBD ++ (4D Video)

PHOTOGRAMMETRY ++ (Static scanning)

* SketchUp>Photogrammetry http://www.mastersketchup.com/why-google-doesnt-need-sketchup-anymore/ * Google Earth https://www.youtube.com/watch?v=N6Douyfa7l8 * Satellite http://eijournal.sensorsandsystems.com/newsite/wp-content/uploads/2012/06/cover_section3_grx.jpg * Matterport http://realestate.matterport.com/ * LIDAR ScanLab http://www.dezeen.com/2014/06/26/scanlab-vivienne-westwood-first-laser-scanned-photoshoot-interview/ * CyArk http://www.cyark.org/education/terrestrial-lidar-3d-laser-scanning * LIDAR for VR http://scanlabprojects.co.uk/projects/newmuseumnewyork * Kinect Fusion https://www.youtube.com/watch?v=of6d7C_ZWwc

Other Techniques

VR: People

VR: Displays

VR: 360-Degree Cameras

VR: Players

Augmented Reality

Unsorted:

tutorials on the photogrammetry work we did on Sunday.

https://vimeo.com/specularprojects/photoscan1 https://vimeo.com/specularprojects/photoscan2

https://www.youtube.com/watch?v=CrsVa3CzQj0 https://vimeo.com/38360045 Snack and Drink

Binocular STEREOGRAPHY

Michael Naimark at Gray Area Festival, 2015

Portraiture Assignment (by James George)

BRIEF

Create an portrait of a subject using a computational photography technique.

CONSTRAINTS

You must choose one of the following explicit constraints when executing your portrait:

  • The viewer of the portrait may interactively influence the perspective of the portrait.
  • The portrait is synthesized from entirely found images that you did not take yourself.
  • Reappropriate a technology that was not intended for expressive photography (i.e. medical imaging, recognition, forensics, etc)
  • The portrait’s final form is not an image, but rather a sculpture.
  • Capture a geometric portrait without a computer, using a manual technique.
  • Create an environmental portrait of a subject using multiple scans composited into a 3D collage.

CONSIDERATIONS

Be mindful of form and content connection. How does your chosen medium or technique help illustrate something unique about your

subject, beyond what a normal photograph could? The environment, effects or geometry should be illustrative of their nature.

Collaborate with and gain the trust of your subject. Make a good faith effort to show them their portrait when it’s complete.

https://vimeo.com/123520067

https://github.com/ivaylopg/RGBDToolkit-to-Processing

http://prostheticknowledge.tumblr.com/post/101297249296/computational-portraiture-a-class-at-nyu-itp-run

KINECT FUSION:

Advantages

  • Fast to process
  • Regular geometry
  • real time feedback
  • non cloud based
  • free tool

Disadvantages

  • limited scale
  • does not work in sunlight
  • color is not very good
  • requires specific sensor with wall power
  • no texture map, per vertex color

123D CATCH

Advantages

  • works at any scale
  • works outside
  • works with any regular camera, even your cell phone
  • generates a texture map
  • Photographic quality in color

Disadvantages

  • Cloud based
  • slow to process
  • dense geometry
  • tools are not free and open
  • no real time feedback
  • more chaotic geometry

https://vimeo.com/109422156 I used Element 3D. It lets you import OBJ files into after effects https://www.videocopilot.net/products/element/ It managed to handle the hi res exported file from 123D scan relatively easy.

I also used Particular to add the floating dust particle ( that you barely see:)) http://www.redgiant.com/products/all/trapcode-particular/

All camera movement is done in After effects

I mentioned that we used both Kinect Fusion and 123D Catch because they are free powerful but that there are other options out there. Here are a few of those tools:

Photogrammetry The paid version -- http://www.agisoft.com/ I used this for the landscapes I showed. It looks like they have just added a free trial! I haven't used it and I don't know the limitations but this tool is a very good and fully featured toolset with a well attended forum. If you want more control than 123D Catch offers I recommend this tool! One highlight with this tool is that you can re-import aligned geometry (remember the Meshlab align tool!) and retexture it using your original photographs.

Open source version(s) -- Visual SFM - http://ccwu.me/vsfm/ http://opensourcephotogrammetry.blogspot.com/ Bundler - http://www.cs.cornell.edu/~snavely/bundler/

Kinect Scanning http://skanect.occipital.com/ http://reconstructme.net/

Last email of the day I promise! remember you can see all the posts here if they are getting lost in your inbox:

https://groups.google.com/forum/#!forum/compportraits

One tool I forgot to mention today is WebGL & Javascript

The Three.js library has native support for both OBJ and PLY and some really nice shaders built in

So if you'd like to use Javascript go for it!

http://threejs.org/examples/webgl_loader_obj.html

http://threejs.org/examples/webgl_loader_ply.html

-- There are also examples out there of people using morph targets (the same as the blendshapes we were looking at today) in three.js.

See the animation examples here: http://threejs.org/examples/#webgl_animation_skinning_blending

https://www.stickmanventures.com/labs/demo/webgl-threejs-morph-target/#

  • show quoted text -

https://vimeo.com/56143282

http://www.makehuman.org/

https://github.com/obviousjim/OFGraphicsITP

OBJ and PLY files are commonly used open 3D file formats. This is what we'll be working with in Computational Portraiture.

The free tool MeshLab provides a very handy way to convert formats. It's like a 3D pocket knife. We'll be using it not just for converting 3D files to different formats, but also reducing and combining meshes. *see below for how to reduce meshes using a few other applications

Here are some more comprehensive tutorials for cleaning geometry in MeshLab https://www.youtube.com/playlist?list=PLBBF41579E4B65566

After capturing your geo, the following links will help you get it into your preferred application

Unity Supports OBJ like any other 3D file format!

Processing The Saito OBJ loader Processing Library https://code.google.com/p/saitoobjloader/ Tutorial Video: https://www.youtube.com/watch?v=6VSaneuiaWs (scrub to 3:30 for the good stuff)

OpenFrameworks The ofMesh and ofVboMesh objects support PLY's with color with the mesh .load() function. This is what you saw in class.

Satoru HIga has a nice OBJ loader addon: https://github.com/satoruhiga/ofxObjLoader

The built in addon ofxAssimpModelLoader also works on OBJ files

Cinder Cinder has built in support for OBJ http://libcinder.org/docs/v0.8.3/classcinder_1_1_obj_loader.html

There is also a Mesh extension block that supports PLY https://github.com/simongeilfus/Cinder-OpenMesh

3D Printing NetFabb will help you clean up models if you wish to 3D print them. You'll want to take them through the STL format in MeshLab http://www.netfabb.com/downloadcenter.php?basic=1

Creative Tools: Many applications support OBJ (and sometimes PLY files)

Photoshop (really!) Photoshop natively supports importing and painting on 3D files! Check out this tutorial https://www.youtube.com/watch?v=mD39wgDoiHE

After Effects The 3D Animator Pro plugin (they offer a free trial) accepts OBJ files https://www.youtube.com/watch?v=eEQUba16V5o

Plexus is an amazing tool for After Effects that works on 3D geometry https://www.youtube.com/watch?v=44hz_fsjX7o

There is also mesh reduction tutorial for the free program Blender https://www.youtube.com/watch?v=ttU6Gz1W0Xw

  • You may also explore MeshMixer and MudBox, which ITP has a student license for through Autodesk

Programs like Maya, Cinema4D, Houdini, 3D Studio Max, Modo also work well with OBJ files but are beyond the scope of what we expect to teach in this class. But if you know these programs, by all means use them for your portraits.

Alignment Meshlab 3D Scanning: Alignment - http://youtu.be/4g9Hap4rX0k

Alignment and Registration in CloudCompare - http://www.danielgm.net/cc/doc/wiki/index.php5?title=Alignment_and_Registration I've found that CloudCompare is really good for loading large pointclouds or found research data in lasercanning formats like .las, .xyz etc. and it will do a great job with Kinect or photogrammetry pointclouds.

Reduction & Remeshing Poisson Reconstruction in Meshlab - This is a good general purpose approach for Kinect or other pointclouds. Can also be used to combine meshes when used after "flatten visible layers" in Meshlab.

Color MeshLab Features: Vertex Attribute Transfer http://youtu.be/sKKmJdsk7Tg?list=UU70CKZQPj_ZAJ0Osrm6TyTg

References from today Kinect Fusion SDK -- http://www.microsoft.com/en-us/kinectforwindows/develop/downloads-docs.aspx http://monoskop.org/images/c/c4/Flusser_Vilem_Towards_a_Philosophy_of_Photography.pdf http://www.creativeapplications.net/tutorials/guide-to-meshes-in-cinder-cinder-tutorials/
  • FaceShift & Unity

Synthetic re-lighting (debevec)

xperimental capture

Full 360: spherical , catadioptric Ricoh theta M15 360 http://www.amazon.com/360-Degree-Spherical-Digital-Monopod-Cleaning/dp/B00TXGYC06/ V.360 http://www.vsnmobil.com/products/v360

Neolucida

Google cardboard,

Microscope Telescope 9dof sensor sparkfun, gps Time lapse Slitscan Panoramic stitching

CanonSDK open frameworks Sensor-Triggered capture 3d point field capture - Larry Sick LIDAR LMS-111 or ...other inexpensive LIDAR Laser distance ranger Flatbed scanner

Fiducials barcodes tuio Opentsps Retroreflective tracker FaceOSC

Sensor tracker: Breath snibbe breath portrait Lozano hammer breath portrait

Visit to CMOA photo archive?

Time lapse panorama http://www.amazon.com/Movo-MTP-10-Motorized-Panaromic-Rechargeable/dp/B00RNALPRC/

View-dependent texturing Multi-camera arrays 2-camera calibration Kinect Depth-Kit slam Point cloud library

Retroreflective sheet

Motion position capture Mouse Phantom Flock of birds Leap controller Eye tribe Kinect skeleton Motion capture lab Rotoscoping Arduino with sensors and datalogger

Cameraless imaging Photograms, lumenprints

Lensless photography Pinhole camera Lytro Kinect

Alan Hauptmann LTI

Cinema4d, Netfabb, cura Scanify by fuel3d, Skanect, autodesk memento

Depth from stereo, 1930s-style: The [Stereoplotter](https://en.wikipedia.org/wiki/Stereoplotter)

Paul Debevec light field capture, reflectance capture

Surveillance camera players Andrew Hammerand 9 eyes of google street view Huge archive of sirvellanxe cameras Motion retargeting BVH reuse mashup Patricio Gonzalez vivo Kyle cat scan Medical imaging Visible human http://www.timothyscaffidi.com/opensource/ofxvolumetrics/ For CT scan http://plum.museum.upenn.edu/~orsa/Overview.html http://www.oasis-brains.org

Faceshift license Wiimote Autodesk motionBuilder (free..) http://www.autodesk.com/education/free-software/motionbuilder

Witrack http://witrack.csail.mit.edu/witrack-paper.pdf

Capture reflectance Scanner - quarter Princeton Debevec Rusinkiewicz Levoy

http://herakles.zcu.cz/~skala/PUBL/PUBL_2013/2013_NormalMapAcquisition.pdf

http://www.zarria.net/nrmphoto/nrmphoto.html ftp://carbonforumasia.com/Public/Melissa/SA2013/A0%20Posters/2%20October/posters_0039.pdf

Norman McLaren “animation is not the art of drawings that move, but rather the art of movements that are drawn”

Zach Lieberman, Inkspace, drawing apps Golan Levin, Yellowtail Merce Cunningham & Open Ended Group Kyle McDonald & YCAM (mocap)

* Nobumichi Asai, [OMOTE 1](https://vimeo.com/103425574) * Nobumichi Asai, [OMOTE 2](https://vimeo.com/117029335)

Kyle & Arturo Face Substitution https://vimeo.com/29348533

GRL Laser Tag https://www.youtube.com/watch?v=cY6vfJAxtH8

The Robotic Cameraman

Come into my world - Gondry

http://imgur.com/LrpYIem selfie stick 900 mpixel portraits https://vimeo.com/94662582?ref=tw-share

Guillaume Reymond human tetris

Perspective camera drawing machines

The Human Clock is a clock made from crowdsourced photographs. The Photo'Clock is similar, by Mono-1.

Michael Kontopoulous, Horizons

SLAM building reconstruction

retroreflection https://www.fbi.gov/about-us/lab/forensic-science-communications/fsc/oct1999/images/coaxialb.jpg