Mercurial > thoughts
view org/adelson-notes.org @ 99:c7272bb5394e
adding more recommendations from loose sheets.
author | Robert McIntyre <rlm@mit.edu> |
---|---|
date | Mon, 26 May 2014 02:19:52 -0400 |
parents | dfcbbb3d4b9a |
children |
line wrap: on
line source
1 #+title: Notes for "Special Topics in Computer Vision"2 #+author: Robert McIntyre3 #+email: rlm@mit.edu4 #+description:5 #+keywords:6 #+SETUPFILE: ../../aurellem/org/setup.org7 #+INCLUDE: ../../aurellem/org/level-0.org8 #+babel: :mkdirp yes :noweb yes :exports both10 * Fri Sep 27 201312 Lambertian surfaces are a special type of Matt surface. They reflect13 light in all directions equally. They have only one parameter, the14 amount of energy that is absorbed/re-emitted.16 [[../images/adelson-checkerboard.jpg]]17 #+caption: Lol checkerboard illusion.19 Look into Helmholtz' stuff, it might be interesting. It was the20 foundation of both vision and audition research. Seems to have took21 a sort of Baysean approach to inferring how vision/audition works.23 - Homomorphic filtering :: Oppenhiem, Schafer, Stockham, 1968. also24 look at Stockham, 1972.26 Edwin Land was Adelson's hero back in the day. He needed to create a27 color photo for the Polaroid camera. In order to process for28 automatic development of film, he had to get a good approximation for29 the illumination/reflectance decomposition that humans do, which he30 called Retinex.32 Cornsweet square wave grating is cool.34 - Retinex :: use derivatives to find illumination. Sort of35 implicitly deals with edges, etc. Can't deal with36 non-lambertian objects.39 Adelson introduces the problem as an "inverse" problem, where you40 try to "undo" the 3-d projection of the world on your retina.42 On the functional view of vision : "What it takes" is to build a43 model of the world in your head. The bare minimum to get success in44 life is to have a model of the world. Even at the level of a single45 cell, I think you still benefit from models.47 Spatial propagation is ABSOLUTELY required to separate embossed48 stuff from "painted" stuff. Edges, likewise, MUST have spatial49 context to disambiguate. The filters we use to deal with edges must50 have larger spatial context to work, and the spatial extent of this51 context must be the ENTIRE visual field in some cases!53 ------------------------------------------------------------55 ** Illumination, shape, reflectance all at once57 What if we tried to infer everything together? Some images are so58 ambiguous it requires propagation from all three qualities to59 resolve the ambiguity.61 Brain has a competing painter, sculptor, and gaffer which each try62 to "build" the things in the world. There is a cost to everything63 such as paints, lights, and material, and then you try to optmize64 some cost function using these primitives.67 Horn, technical report, 197069 * Fri Oct 4 201371 Student report. Talked about how you capture the appearance of a72 grape. It's actually quite compicated, involving gloss, spatial73 context, etc.75 Turbosquid seems interesting. They sell 3D models of stuff.77 BRDF -- bi-directional reflectance distribution function this shows78 how a surface will behave given lighting conditions. Lambertian is a79 simple parameterized instantiation of this.81 BSSRDF -- (SS = subsurface) 3D analogue of BRDF83 What would the 3D analiogue of texture be?85 (a : b : c) as (a + b + c : b + c : c) <-- this is just the golden86 ratio again!88 CURET BTF Database lol what's this90 This student went and gathered 1000 images of different large91 objects made of different materials. The images were gathered off of92 Flikr.94 Then she gave another talk from someone else. It's about assigning95 materials to objects and then rendering them. The choice of96 materials is determined by some sort of expert system?98 They have made a neat looking interface for human entry of texture99 labeling or objects in scenes. The important elements were manual100 labels, dynamic display of the current selection, and undo.102 There are papers about mechanical turk engineering.104 lol http://opensurfaces.cs.cornell.edu/106 CUBAM is interesting 2010 "The multidimensional wisdom of crowds."107 "Neural Information processing systems" some sort of voting scheme.109 The point of this is apparently to do some kitchen makeover110 thing. You would take a picture of your kitchen, or you would look111 for a kitchen that looks like yours, and then you would be able to112 investigate different textures for your own kitchen.114 apparently "Label ME" has never been appropriately115 crowdsourced. Turns out that you get better work if you don't use116 idiots to do the work lol.