Interactive Pattern Recognition

Vidal, Enrique; Rodríguez, Luis; Casacuberta, Francisco; García-Varea, Ismael

doi:10.1007/978-3-540-78155-4_6

Cited by 37 publications

(30 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Insofar as human experience enters the picture, it is merely in the form of behavioral patterns which are modeled-in supervised learning, interactive segmentation, activity recognition etc. While there are exceptions to this stereotyping [32,33], we feel our generalization is a fair characterization. We argue that this lack is due to the influence of strong AI which is still being felt by these fields long after AI has lost its once dominant position.…”

Section: Twenty First Century Schizoid Manmentioning

confidence: 71%

Revisioning the unification of syntax, semantics and statistics in shape analysis

Rangarajan

2014

Pattern Recognition Letters

View full text Add to dashboard Cite

Hidden patterns, rendered visible by hindsight, often stand revealed as strong influences. Rama Chellappa's lab in the mid to late '80s molded our character and scholarship in more ways than one. Rama ably handled the transition from image processing to computer vision and established an applied math and computing infrastructure from which we continue to benefit. In particular, the themes important to King-Sun Fu-syntax, semantics and statistics-were all debated in Rama's lab at that time. We argue that this triad remains important. With syntactic representations losing mindshare to statistics, we remain in the hunt for unification. And with the syntax versus semantics debate unresolved, it deserves a hearing as well. We offer two themes-uncertainty and interaction-to aid in the process of unification. First, we show that complex wave functions carry probabilistic location information in their magnitude and syntactic (curve) information in their phase while representing uncertainty at a fundamental level. Next, after reviewing work in analytic philosophy, we connect semantics to intentional, mental content. Analytic philosophy reminds us to take human experience seriously but remaining physicalist if possible. To this end, we introduce a nondualist interactionist model of experience, wherein compositional (physical) subjects are constantly shaping and being shaped by a physical world. We then demonstrate that wave functions can accommodate * Corresponding Author Email address: anand@cise.ufl.edu (Anand Rangarajan) 1 This work is partially supported by NSF IIS RI:1143963. Preprint submitted to Pattern Recognition LettersJuly 22, 2013 interaction, closely tracking previous work in physics on the measurement problem. The linearity and superposition properties of wave functions allow for literal addition of waves created by human interaction with shapes. Finally, we briefly survey the current situation in the human-computer interaction (HCI) field and argue that mathematical models of interaction akin to those in pattern recognition can aid HCI. We close by arguing that we can follow in Fu's footsteps and incorporate the mathematical modeling of human interaction into pattern recognition.Keywords: syntax, semantics, statistics, uncertainty, interaction, wave functions, Schrödinger, Hamilton-Jacobi, phase, topology, level-sets, distance transform, intentionality, qualia, compositionality, HCI InvocationHold the flame 'til the dream ignites A spirit with a vision is a dream with a mission Rush, MissionLuck is often described as "being in the right place at the right time". Looking back, it seems clear that we were very fortunate indeed to inhabit Prof. While we return to these themes frequently in this work, it is to the specifics of Rama's contributions-and the manner in which they illustrate the deployment of applied math in computer vision-that we now turn.Since we're using a signal processing to computer vision crossover perspective, we bypass Rama's considerable work on Markov random fields in image pr...

show abstract

Section: Twenty First Century Schizoid Manmentioning

confidence: 71%

Revisioning the unification of syntax, semantics and statistics in shape analysis

Rangarajan

2014

Pattern Recognition Letters

View full text Add to dashboard Cite

show abstract

“…In this paper we are interested in applying Interactive PR (IPR) [11] approaches to problems with structured output because it is in this kind of problems where the IPR framework is likly to be most fruitful.…”

Section: Introductionmentioning

confidence: 99%

Interactive Structured Output Prediction: Application to Chromosome Classification

Oncina

Vidal

2011

Pattern Recognition and Image Analysis

Self Cite

View full text Add to dashboard Cite

Abstract. Interactive Pattern Recognition concepts and techniques are applied to problems with structured output; i.e., problems in which the result is not just a simple class label, but a suitable structure of labels. For illustration purposes (a simplification of) the problem of Human Karyotyping is considered. Results show that a) taking into account label dependencies in a karyogram significantly reduces the classical (noninteractive) chromosome label prediction error rate and b) they are further improved when interactive processing is adopted.

show abstract

“…-Average number of corrections (C): a good way in which to measure the performance of this kind of systems is to count the number of corrections the user would have to make before obtaining the sequence actually written (Vidal et al, 2008). After each correction, the system is allowed to recompute a new hypothesis, which is expected to be more accurate since 490 some parts of the solution are known.…”

mentioning

confidence: 99%

Recognition of pen-based music notation with finite-state machines

Calvo-Zaragoza

Oncina

2017

Expert Systems with Applications

View full text Add to dashboard Cite

This work presents a data-driven statistical approach with which to recognize pen-based music compositions. Unlike previous works, no assumption is made as regards the handwriting notation, but the system is able to adapt to any kind of style obtained from training data. The series of strokes received as input is mapped onto a stochastic representation, which is combined with a formal language that describes musical symbols in terms of stroke primitives.A Probabilistic Finite-State Automaton is then obtained, which defines probabilities over the set of musical sequences. This model is eventually crossed with a semantic language to avoid sequences that do not make musical sense. Finally, a decoding strategy is applied in order to output a hypothesis concerning the musical sequence actually written. Comprehensive experimentation with several decoding algorithms, stroke similarity measures and probability density estimators are tested and evaluated following different metrics of interest to both user-dependent and user-independent scenarios. The results demonstrate that it is possible to learn the handwriting music style, obtaining competitive performances in each case considered.

show abstract

Interactive Pattern Recognition

Cited by 37 publications

References 7 publications

Revisioning the unification of syntax, semantics and statistics in shape analysis

Revisioning the unification of syntax, semantics and statistics in shape analysis

Interactive Structured Output Prediction: Application to Chromosome Classification

Recognition of pen-based music notation with finite-state machines

Contact Info

Product

Resources

About