Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
Â
Fcv scene efros
1. What do we want from
Computational Scene Understanding?
Š Quint Buchholz
Alexei âAlyoshaâ Efros
Carnegie Mellon University
2. Many ways to Understand a SceneâŚ
⢠Scene Categories (e.g. street scene, city, Boston, etc.)
⢠Innumerate objects (people, cars, lampposts, etc)
⢠Label/Segment scene elements (road, buildings, sky, etc)
⢠Scene Geometry (qualitative or quantitative) and Illumination
⢠Physical Affordances (where can I walk?)
⢠Prediction (What will happen next?)
3. Scene Categorization
beach mountain forest
Animal vs. no animal (Thorpe, city street farm
Poggio & Oliva)
Basic Scene Categories (Oliva, Renenger, Fei-Fei, etc)n
Image classification, e.g. âBoatâ (Caltech, etc.)
Spatial envelope (Oliva & Torralba) im2gps
4. Poster Spotlight: SUN Attributes: A Large-Scale
Database of Scene Attributes
Genevieve Patterson and James Hays, Brown University
Global, binary attributes
describing:
⢠Affordances / Functions (e.g. farming,
eating)
⢠Materials (e.g. carpet, running water)
⢠Surface Properties (e.g. aged, sterile)
⢠Spatial Envelope (e.g. enclosed,
symmetrical)
Statistics of database:
⢠14340 images from 717 scene
categories
⢠106 attributes
Space of Scenes ⢠2 million+ labels collected so far
Organized by Attributes ⢠Outlier workers manually graded,
good workers ~90% accurate.
11. Pushing Back Evaluation HorizonâŚ
⢠So far, we have acted as cognitive
psychologists:
â proposing and evaluating intermediate âmental
representationsâ
â E.g. object/scene categories, pixel labels,
geometry
⢠But we can also be more behaviorist:
â Focus on tasks instead of âmental statesâ
â Evaluate action plans and behavior predictions
19. What do we need
to get to there?
The Op-Ed Part âş
20. Organizing Our Data
âIt irritated him that the âdogâ of 3:14 in the
afternoon, seen in profile, should be
indicated by the same noun as the dog of
3:15, seen frontallyâŚâ
âMy memory, sir, is like a garbage heap.â
Fumes the Memorious
Jorge Luis Borges
21. Trouble with Classic Platonic
Categorization
⢠Step 1: cut up the world into
categories
PASCAL
âtrainâ
category
⢠Step 2: train an SVM on
positives vs. negatives
22. It gets more complicatedâŚ
⢠Number of objects * number of interactions *
number of outcomes⌠= too many categories
⢠Donât want to categorize too early
â âDealing with the world as it comes to usâ [Derek]
⢠Letâs categorize at run-time, once we know the task!
23. The Dictatorship of Librarians
Arts and recreation
Arts and recreation Language
Language
Philosophy and
Philosophy and
Literature
Literature Psychology
Psychology
Technology
Technology Religion
Religion
23
25. Association instead of
categorization
Ask not âwhat is this?â, ask âwhat is this likeâ
â Moshe Bar
⢠Exemplar Theory (Medin & Schaffer 1978,
Nosofsky 1986, Krushke 1992)
âcategories represented in terms of remembered objects
(exemplars)
âSimilarity is measured between input and all exemplars
âthink non-parametric density estimation
⢠Vanevar Bush (1945), Memex (MEMory
EXtender)
âInspired hypertext, WWW, GoogleâŚ
26. Bushâs Memex (1945)
⢠Store publications, correspondence, personal work, on microfilm
⢠Items retrieved rapidly using index codes
â Builds on ârapid selectorâ
⢠Can annotate text with margin notes, comments
⢠Can construct a trail through the material and save it
â Roots of hypertext
⢠Acts as an external memory
28. Poster Spotlight: Relative Attributes
⢠Previous work restricts attributes to binary categories,
but many attributes are more fluid and should be
expressed relatively.
[Parikh & Grauman, ICCV 2011]
29. Relative attributes
[Parikh & Grauman, ICCV 2011]
⢠Learn a ranking function per attribute, given ordering
constraints among exemplars or categories
Youth:
, âŚ
⢠Allows two novel tasks:
1) Zero-shot learning from 2) Description relative to
comparisons examples/classes
Train: âUnseen person C is
younger than S, older than Hâ,âŚ
is more dense than ,
S
Smiling
and less dense than
C M
H Z Precise descriptions are more
recognizable to human subjects
Youth
30. Poster Spotlight: Ensemble of Exemplar-SVMs
for Object Detection and Beyond
Milosewicz,Gupta,Efros, ICCV 2011]