Title: Lectureship Early Career Fellowship
1LectureshipEarly Career Fellowship
Fabio Cuzzolin INRIA Rhone-Alpes
- School of Technology, Oxford Brookes University
- 19/6/2008
2Career path
- Masters thesis on gesture recognition at the
University of Padova - Visiting student, ESSRL, Washington University in
St. Louis, and at the University of California at
Los Angeles (2000) - Ph.D. thesis on belief functions and uncertainty
theory (2001)? - Researcher at Politecnico di Milano with the
Image and Sound Processing group (2003-2004)? - Post-doc at the University of California at Los
Angeles, UCLA Vision Lab (2004-2006)? - Marie Curie fellow at INRIA Rhone-Alpes
3Scientific production and collaborations
- collaborations with journals
IEEE PAMI
IEEE SMC-B
CVIU
Information Fusion
Int. J. Approximate Reasoning
- PC member for VISAPP, FLAIRS, IMMERSCOM, ISAIM
- currently 410 journal papers and 318 conference
papers
4My background
research
5Computer Vision
- Action and gesture recognition
- Laplacian segmentation and matching of 3D shapes
- Bilinear models for invariant gaitID
Machine learning
Manifold learning for dynamical models
Uncertainty theory
Geometric approach to measures
Generalized total probability Vision applications
and developments
Discrete math
Unification of the notion of independence
New directions
6HMMs for gesture recognition
- transition matrix A -gt gesture dynamics
- state-output matrix C -gt collection of hand poses
- Hand poses were represented by size functions
(BMVC'97)?
7Gesture classification
- EM to learn HMM parameters from an input sequence
- the new sequence is fed to the
- learnt gesture models
- they produce a likelihood
- the most likely model is chosen (if above a
threshold)? - OR new model is attributed the label of the
closest one (using K-L divergence or other
distances)?
HMM 1
HMM 2
HMM n
8Compositional behavior of HMMs
- the model of the action of interest is embedded
in the overall model ? clustering
- Cluttered model for two
- overlapping motions
- Reduced model for the fly gesture after
clustering
9Volumetric action recognition
- 2D approaches, feature extracted from images ?
viewpoint dependence - now available synchronized multi-camera systems
Milano, BBC RD - volumetric approach features are extracted from
volumetric reconstructions of the body (ICIP'04)?
103D feature extraction
- k-means clustering of bodyparts
- Linear discriminant analysis (LDA) to estimate
motion direction
- Locally linear embedding to find topological
representation of the moving body
11Computer Vision
- Action and gesture recognition
- Laplacian segmentation and matching of 3D shapes
- Bilinear models for invariant gaitID
Machine learning
Manifold learning for dynamical models
Uncertainty theory
Geometric approach to measures
Generalized total probability Vision applications
and developments
Discrete math
Unification of the notion of independence
New directions
12Unsupervised coherent 3D segmentation
- to recognize actions we need to extract features
- segmenting moving articulated 3D bodies into
parts - along sequences, in a consistent way
- in an unsupervised fashion
- robustly, with respect to changes of the topology
of the moving body - as a building block of a wider motion analysis
and capture framework - ICCV-HM'07, CVPR'08, to submit to IJCV
13Clustering after Laplacian embedding
- local neighbourhoods stable under articulated
motion
- generates a lower-dim, widely separated embedded
cloud - less sensitive to topology changes than other
methods - less expensive then ISOMAP (refs. Jenkins,
Chellappa)?
14Algorithm
- K-wise clustering in the embedding space
15Seed propagation along time
- To ensure time consistency clusters seeds have
to be propagated along time - Old positions of clusters in 3D are added to new
cloud and embedded - Result new seeds
16Results
- Example of model recovery
- Coherent clustering along a sequence
17Results - 2
- handling of topology changes
18Laplacian matching of dense meshes or voxelsets
- as embeddings are pose-invariant (for articulated
bodies)? - they can then be used to match dense shapes by
simply aligning their images after embedding
- ICCV '07 NTRL, ICCV '07 3dRR, CVPR '08,
submitted to ECCV'08, to submit to PAMI
19Eigenfunction Histogram assignment
- Algorithm
- compute Laplacian embedding of the two shapes
- find assignment between eigenfunctions of the two
shapes - this selects a section of the embedding space
- embeddings are orthogonally aligned there by EM
20Results
- Appls graph matching, protein analysis, motion
capture - To propagate bodypart segmentation in time
- Motion field estimation, action segmentation
21Computer Vision
- Action and gesture recognition
- Laplacian segmentation and matching of 3D shapes
- Bilinear models for invariant gaitID
Machine learning
Manifold learning for dynamical models
Uncertainty theory
Geometric approach to measures
Generalized total probability Vision applications
and developments
Discrete math
Unification of the notion of independence
New directions
22Bilinear models for gait-ID
- To recognize the identity of humans from their
gait (CVPR '06, book chapter in progress)? - nuisance factors emotional state, illumination,
appearance, view invariance ... (literature
randomized trees)?? - each motion possess several labels action,
identity, viewpoint, emotional state, etc.
- bilinear models Tenenbaum can be used to
separate the influence of style and content
(to classify)?
23Content classification of unknown style
- given a training set in which persons
(contentID) are seen walking from different
viewpoints (styleviewpoint)? - an asymmetric bilinear model can learned from it
through SVD - when new motions are acquired in which a known
person is being seen walking from a different
viewpoint (unknown style) - an iterative EM procedure can be set up to
classify the content - E step -gt estimation of p(cs), the prob. of the
content given the current estimate s of the style
- M step -gt estimation of the linear map for
unknown style s
24Three-layer model
- Features projections of silhouette's contours
onto a line through the center
- Three layer model
- each sequence is encoded as an HMM
- its C matrix is stacked in a single observation
vector - a bilinear model is learnt from those vectors
25Results on CMU database
- Mobo database 25 people performing 4 different
walking actions, from 6 cameras. Three labels
action, id, view - Compared performances with baseline algorithm
and straight k-NN on sequence HMMs
26Computer Vision
- Action and gesture recognition
- Laplacian segmentation and matching of 3D shapes
- Bilinear models for invariant gaitID
Machine learning
Manifold learning for dynamical models
Uncertainty theory
Geometric approach to measures
Generalized total probability Vision applications
and developments
Discrete math
Unification of the notion of independence
New directions
27Learning manifolds of dynamical models
- Classify movements represented as dynamical
models - for instance, each image sequence can be mapped
to an ARMA, or AR linear model, or a HMM - Motion classification then reduces to find a
suitable distance function in the space of
dynamical models - e.g. Kullback-Leibler, Fisher metric Amari
- when some a-priori info is available (training
set).. - .. we can learn in a supervised fashion the
best metric for the classification problem! - To submit to ECCV'08 MLVMA Workshop
28Learning pullback metrics
- many algorithms take in input dataset and map it
to an embedded space, but fail to learn a full
metric (LLE, ISOMAP)? - consider than a family of diffeomorphisms F?
between the original space M and a metric space N - the diffeomorphism F induces on M a pullback
metric - maximizing inverse volume finds the manifold
which better interpolates the data (geodesics
pass through crowded regions)?
29Pullback metrics - detail
- case of linear maps Xing and Jordan'02,
Shental'02
30Space of AR(2) models
- given an input sequence, we can identify the
parameters of the linear model which better
describes it - autoregressive models of order 2 AR(2)?
- Fisher metric on AR(2)?
- Compute the geodesics of the pullback metric on M
31Results on action and ID rec
- scalar feature, AR(2) and ARMA models
32Computer Vision
- Action and gesture recognition
- Laplacian segmentation and matching of 3D shapes
- Bilinear models for invariant gaitID
Machine learning
Manifold learning for dynamical models
Uncertainty theory
Geometric approach to measures
Generalized total probability Vision applications
and developments
Discrete math
Unification of the notion of independence
New directions
33Uncertainty measures Intervals, credal sets
- a number of formalisms have been proposed to
extend or replace classical probability
- assumption not enough evidence to determine the
actual probability describing the problem - second-order distributions (Dirichlet), interval
probabilities - credal sets
34Multi-valued maps and belief functions
- suppose you have two different but related
problems ... - ... that we have a probability distribution for
the first one - ... and that the two are linked by a map one to
many - Dempster'68, Shafer'76
35Belief functions as random sets
- probability on a finite set function p 2T -gt
0,1 with - p(A)?x e A m(x), where m T -gt 0,1 is a
mass function
- probabilities are additive if A?B? then
p(A?B)p(A)p(B)?
36Geometric approach to uncertainty
- belief functions can be seen as points of a
Cartesian space of dimension 2n-2 - belief space the space of all the belief
functions on a given frame
- Each subset is a coordinate in this space
37 Approximation problem
- how to transform a measure of a certain family
into a different uncertainty measure ? can be
done geometrically
- Probabilities, fuzzy sets, possibilities are
special cases of b.f.s
- IEEE Tr. SMC-B '07, IEEE Tr. Fuzzy Systems '07,
AMAI '08, AI '08, IEEE Tr. Fuzzy Systems '08
38Computer Vision
- Action and gesture recognition
- Laplacian segmentation and matching of 3D shapes
- Bilinear models for invariant gaitID
Machine learning
Manifold learning for dynamical models
Uncertainty theory
Geometric approach to measures
Generalized total probability Vision applications
and developments
Discrete math
Unification of the notion of independence
New directions
39Total belief theorem
- generalization of the total probability theorem
- introduces Kalman-like filtering for random sets
40Graph of all solutions
- whole graph of candidate solutions
- admissible solution is found by following a path
on the graph - links to combinatorics and linear systems
- to submit to JRSS-B
41Computer Vision
- Action and gesture recognition
- Laplacian segmentation and matching of 3D shapes
- Bilinear models for invariant gaitID
Machine learning
Manifold learning for dynamical models
Uncertainty theory
Geometric approach to measures
Generalized total probability Vision applications
and developments
Discrete math
Unification of the notion of independence
New directions
42Model-free pose estimation
- estimating the pose (internal configuration) of
a moving body from the available images
- if you do not have an a-priori model of the
object .. - Sun Torr BMVC'06, Rosales, Urtasun Brand,
Grauman ICCV'03, Agarwal
43Learning feature-pose maps
- ... learn a map between features and poses
directly from the data - given pose and feature sequences acquired by
motion capture ..
- a multi-modal Gaussian density is set up on
the feature space - a map from each cluster to the set of poses
whose feature values fall inside it (regression
functions, EM)?
44Evidential model
- similar to propagation in qualitative Markov
trees - MTNS'00, ISIPTA'05, to submit to Information
Fusion
45Information fusion by Dempsters rule
- several aggregation or elicitation operators
proposed - original proposal Dempsters rule
- b2
- m(?)0.1, m(a2 ,a3 ,a4)0.9
46Performances
- comparison of three models left view only, right
view only, both views
- estimate associated with the right model
- pose estimation yielded by the overall model
47JPDA with shape info
- JPDA model independent targets
- robustness clutter does not meet shape
constraints - occlusions occluded targets can be estimated
- CDC'02, CDC'04
48Belief graphical models
- what happens when the original probability
distribution belongs to a certain class? - In particular belief functions induced by
graphical models?
49Imprecise classifiers
- application of robust statistics to vision
problems - imprecise classifiers
- class estimate is a belief function or a credal
set Zaffalon, Cozman - exploit only available evidence, represent
ignorance
50Credal networks
- belief networks or credal networks Shafer and
Shenoy - at each node a BF or a convex set of probs
- similar to generalized belief propagation ...
- message passing between nodes representing groups
of variables
- algorithms to reduce complexity already exist
51Computer Vision
- Action and gesture recognition
- Laplacian segmentation and matching of 3D shapes
- Bilinear models for invariant gaitID
Machine learning
Manifold learning for dynamical models
Uncertainty theory
Geometric approach to measures
Generalized total probability Vision applications
and developments
Discrete math
Unification of the notion of independence
New directions
52 Boolean independence
- independence can be defined in different ways in
Boolean algebras, semi-modular lattices, and
matroids - Boolean independence is important in uncertainty
theory
- a set of sub-algebras At of a Boolean algebra B
are independent (IB) if
- example collection of power sets of the
partitions of a given finite set
53Relation with matroids?
- Matroids ? paradigm of abstract independence
- matroid (E, I?2E)
- ??I
- A?I, A?A then A?I
- A1?I, A2?I, A2gtA1 then ?x ? A2 s.t. A1?x?I
- graphic matroids dependent sets are circuits
- they have significant relationships BUT
- Boolean independence a form of
anti-matroidicity? - BCC'01, BCC'07, ISAIM'08, UNCLOG'08, subm.to
Discrete Mathematics
54Computer Vision
- Action and gesture recognition
- Laplacian segmentation and matching of 3D shapes
- Bilinear models for invariant gaitID
Machine learning
Manifold learning for dynamical models
Uncertainty theory
Geometric approach to measures
Generalized total probability Vision applications
and developments
Discrete math
Unification of the notion of independence
New directions
55A multi-layer frameworkfor human motion analysis
- feedbacks act between different layers (e.g.
integrated detection, segmentation,
classification and pose estimation)?
56Spatio-temporal action segmentation
- problem segmenting parts of the video(s)
containing interesting motions - global approach working on
multidimensional volumes - previous works object segmentation
on the
spatio-temporal volume for
single frames Collins,
Natarajan - idea in a multi-camera setup, working on 3D
clouds (hulls) motion fields time 7D volume - proposal smoothing using tensor voting Medioni
PAMI'05 shape detection on the obtained
manifold
57Stereo correspondence based on local image
structure
- problem finding correspondences between points
in different view, using the local structure of
the image - Markov random fields disparity hidden variable
- one direction using local direction of the
gradient or structure tensor to help the
correspondence Zucker - second option FRAME -gt large scale structures in
MRF - general potential for MRFs, local texture for
correspondence?
58Other developments
- 3D markerless motion capture
- Proposal data-driven pose estimation based on 3D
representations - unsupervised body model learning
- shape classification/ recognition in embedding
spaces - surveillance in crowded areas impossible to
recover a 3D model - ? information fusion techniques on multiple
images - handle conflict between different pieces of
evidence