Learning Embeddings for Indexing, Retrieval, and Classification, with Applications to Object and Shape Recognition in Image Databases

Nearest neighbor retrieval is the task of identifying, given a database of objects and a query object, the objects in the database that are the most similar to the query. Retrieving nearest neighbors is a necessary component of many practical applications, in fields as diverse as computer vision, pattern recognition, multimedia databases, bioinformatics, and computer networks. At the same time, finding nearest neighbors accurately and efficiently can be challenging, especially when the database contains a large number of objects, and when the underlying distance measure is computationally expensive. This thesis proposes new methods for improving the efficiency and accuracy of nearest neighbor retrieval and classification in spaces with computationally expensive distance measures. The proposed methods are domain-independent, and can be applied in arbitrary spaces, including non-Euclidean and non-metric spaces. In this thesis particular emphasis is given to computer vision applications related to object and shape recognition, where expensive non-Euclidean distance measures are often needed to achieve high accuracy. The first contribution of this thesis is the BoostMap algorithm for embedding arbitrary spaces into a vector space with a computationally efficient distance measure. Using this approach, an approximate set of nearest neighbors can be retrieved efficiently - often orders of magnitude faster than retrieval using the exact distance measure in the original space. The BoostMap algorithm has two key distinguishing features with respect to existing embedding methods. First, embedding construction explicitly maximizes the amount of nearest neighbor information preserved by the embedding. Second, embedding construction is treated as a machine learning problem, in contrast to existing methods that are based on geometric considerations. The second contribution is a method for constructing query-sensitive distance measures for the purposes of nearest neighbor retrieval and classification. In high-dimensional spaces, query-sensitive distance measures allow for automatic selection of the dimensions that are the most informative for each specific query object. It is shown theoretically and experimentally that query-sensitivity increases the modeling power of embeddings, allowing embeddings to capture a larger amount of the nearest neighbor structure of the original space. The third contribution is a method for speeding up nearest neighbor classification by combining multiple embedding-based nearest neighbor classifiers in a cascade. In a cascade, computationally efficient classifiers are used to quickly classify easy cases, and classifiers that are more computationally expensive and also more accurate are only applied to objects that are harder to classify. An interesting property of the proposed cascade method is that, under certain conditions, classification time actually decreases as the size of the database increases, a behavior that is in stark contrast to the behavior of typical nearest neighbor classification systems. The proposed methods are evaluated experimentally in several different applications: hand shape recognition, off-line character recognition, online character recognition, and efficient retrieval of time series. In all datasets, the proposed methods lead to significant improvements in accuracy and efficiency compared to existing state-of-the-art methods. In some datasets, the general-purpose methods introduced in this thesis even outperform domain-specific methods that have been custom-designed for such datasets.

How Does the Cerebral Cortex Work? Developement, Learning, Attention, and 3D Vision by Laminar Circuits of Visual Cortex

A key goal of behavioral and cognitive neuroscience is to link brain mechanisms to behavioral functions. The present article describes recent progress towards explaining how the visual cortex sees. Visual cortex, like many parts of perceptual and cognitive neocortex, is organized into six main layers of cells, as well as characteristic sub-lamina. Here it is proposed how these layered circuits help to realize the processes of developement, learning, perceptual grouping, attention, and 3D vision through a combination of bottom-up, horizontal, and top-down interactions. A key theme is that the mechanisms which enable developement and learning to occur in a stable way imply properties of adult behavior. These results thus begin to unify three fields: infant cortical developement, adult cortical neurophysiology and anatomy, and adult visual perception. The identified cortical mechanisms promise to generalize to explain how other perceptual and cognitive processes work.

A Quantitative Evaluation of the AVITEWRITE Model of Handwriting Learning

Much sensory-motor behavior develops through imitation, as during the learning of handwriting by children. Such complex sequential acts are broken down into distinct motor control synergies, or muscle groups, whose activities overlap in time to generate continuous, curved movements that obey an intense relation between curvature and speed. The Adaptive Vector Integration to Endpoint (AVITEWRITE) model of Grossberg and Paine (2000) proposed how such complex movements may be learned through attentive imitation. The model suggest how frontal, parietal, and motor cortical mechanisms, such as difference vector encoding, under volitional control from the basal ganglia, interact with adaptively-timed, predictive cerebellar learning during movement imitation and predictive performance. Key psycophysical and neural data about learning to make curved movements were simulated, including a decrease in writing time as learning progresses; generation of unimodal, bell-shaped velocity profiles for each movement synergy; size scaling with isochrony, and speed scaling with preservation of the letter shape and the shapes of the velocity profiles; an inverse relation between curvature and tangential velocity; and a Two-Thirds Power Law relation between angular velocity and curvature. However, the model learned from letter trajectories of only one subject, and only qualitative kinematic comparisons were made with previously published human data. The present work describes a quantitative test of AVITEWRITE through direct comparison of a corpus of human handwriting data with the model's performance when it learns by tracing human trajectories. The results show that model performance was variable across subjects, with an average correlation between the model and human data of 89+/-10%. The present data from simulations using the AVITEWRITE model highlight some of its strengths while focusing attention on areas, such as novel shape learning in children, where all models of handwriting and learning of other complex sensory-motor skills would benefit from further research.

Learning and Production of Movement Sequences: Behavioral, Neurophysiological, and Modeling Perspectives

A growing wave of behavioral studies, using a wide variety of paradigms that were introduced or greatly refined in recent years, has generated a new wealth of parametric observations about serial order behavior. What was a mere trickle of neurophysiological studies has grown to a more steady stream of probes of neural sites and mechanisms underlying sequential behavior. Moreover, simulation models of serial behavior generation have begun to open a channel to link cellular dynamics with cognitive and behavioral dynamics. Here we summarize the major results from prominent sequence learning and performance tasks, namely immediate serial recall, typing, 2XN, discrete sequence production, and serial reaction time. These populate a continuum from higher to lower degrees of internal control of sequential organization. The main movement classes covered are speech and keypressing, both involving small amplitude movements that are very amenable to parametric study. A brief synopsis of classes of serial order models, vis-à-vis the detailing of major effects found in the behavioral data, leads to a focus on competitive queuing (CQ) models. Recently, the many behavioral predictive successes of CQ models have been joined by successful prediction of distinctively patterend electrophysiological recordings in prefrontal cortex, wherein parallel activation dynamics of multiple neural ensembles strikingly matches the parallel dynamics predicted by CQ theory. An extended CQ simulation model-the N-STREAMS neural network model-is then examined to highlight issues in ongoing attemptes to accomodate a broader range of behavioral and neurophysiological data within a CQ-consistent theory. Important contemporary issues such as the nature of working memory representations for sequential behavior, and the development and role of chunks in hierarchial control are prominent throughout.

A Dopamine-Acetylcholine Cascade: Simulating Learned and Lesion-Induced Behavior ff Striatal Cholinergic Interneurons

The "teaching signal" that modulates reinforcement learning at cortico-striatal synapses may be a sequence composed of an adaptively scaled DA burst, a brief ACh burst, and a scaled ACh pause. Such an interpretation is consistent with recent data on cholinergic interneurons of the striatum are tonically active neurons (TANs) that respond with characteristic pauses to novel events and to appetitive and aversive conditioned stimuli. Fluctuations in acetylcholine release by TANs modulate performance- and learning- related dynamics in the striatum. Whereas tonic activity emerges from intrinsic properties of these neurons, glutamatergic inputs from thalamic centromedian-parafascicular nuclei, and dopaminergic inputs from midbrain are required for the generation of pause responses. No prior computational models encompass both intrinsic and synaptically-gated dynamics. We present a mathematical model that robustly accounts for behavior-related electrophysiological properties of TANs in terms of their intrinsic physiological properties and known afferents. In the model balanced intrinsic hyperpolarizing and depolarizing currents engender tonic firing, and glutamatergic inputs from thalamus (and cortex) both directly excite and indirectly inhibit TANs. If the latter inhibition, probably mediated by GABAergic NOS interneurons, exceeds a threshold, its effect is amplified by a KIR current to generate a prolongued pause. In the model, the intrinsic mechanisms and external inputs are both modulated by learning-dependent dopamine (DA) signals and our simulations revealed that many learning-dependent behaviors of TANs are explicable without recourse to learning-dependent changes in synapses onto TANs.

Simulating Effects of Learning and Lesions with a Model of Intrinsic and Synaptically Gated Responses of Striatal Cholinergic Interneurons

The giant cholinergic interneurons of the striatum are tonically active neurons (TANs) that respond with characteristic pauses to novel events and to appetitive and aversive conditioned stimuli. Fluctuations in acetylcholine release by TANs modulate performance- and learning-related dynamics in the striatum. Whereas tonic activity emerges from intrinsic properties of these neurons, glutamatergic inputs from thalamic centromedian-parafascicular nuclei, and dopaminergic inputs from midbrain, are required for the generation of pause responses. No prior computational models encompass both intrinsic and synaptically-gated dynamics. We present a mathematical model that robustly accounts for behavior-related electrophysiological properties of TANs in terms of their intrinsic physiological properties and known afferents. In the model, balanced intrinsic hyperpolarizing and depolarizing currents engender tonic firing, and glutamatergic inputs from thalamus (and cortex) both directly excite and indirectly inhibit TANs. If the latter inhibition, presumably mediated by GABAergic interneurons, exceeds a threshold, its effect is amplified by a KIR current to generate a prolonged pause. In the model, the intrinsic mechanisms and external inputs are both modulated by learning-dependent dopamine (DA) signals and our simulations revealed that many learning-dependent behaviors of TANs are explicable without recourse to learning-dependent changes in synapses onto TANs. The "teaching signal" that modulates reinforcement learning at cortico-striatal synapses may be a sequence composed of an adaptively scaled DA burst, a brief ACh burst, and a scaled ACh pause. Such an interpretation is consistent with recent data on cholinergic control of LTD of cortical synapses onto striatal spiny projection neurons.

Vector Associative Maps: Unsupervised Real-time Error-based Learning and Control of Movement Trajectories

This article describes neural network models for adaptive control of arm movement trajectories during visually guided reaching and, more generally, a framework for unsupervised real-time error-based learning. The models clarify how a child, or untrained robot, can learn to reach for objects that it sees. Piaget has provided basic insights with his concept of a circular reaction: As an infant makes internally generated movements of its hand, the eyes automatically follow this motion. A transformation is learned between the visual representation of hand position and the motor representation of hand position. Learning of this transformation eventually enables the child to accurately reach for visually detected targets. Grossberg and Kuperstein have shown how the eye movement system can use visual error signals to correct movement parameters via cerebellar learning. Here it is shown how endogenously generated arm movements lead to adaptive tuning of arm control parameters. These movements also activate the target position representations that are used to learn the visuo-motor transformation that controls visually guided reaching. The AVITE model presented here is an adaptive neural circuit based on the Vector Integration to Endpoint (VITE) model for arm and speech trajectory generation of Bullock and Grossberg. In the VITE model, a Target Position Command (TPC) represents the location of the desired target. The Present Position Command (PPC) encodes the present hand-arm configuration. The Difference Vector (DV) population continuously.computes the difference between the PPC and the TPC. A speed-controlling GO signal multiplies DV output. The PPC integrates the (DV)·(GO) product and generates an outflow command to the arm. Integration at the PPC continues at a rate dependent on GO signal size until the DV reaches zero, at which time the PPC equals the TPC. The AVITE model explains how self-consistent TPC and PPC coordinates are autonomously generated and learned. Learning of AVITE parameters is regulated by activation of a self-regulating Endogenous Random Generator (ERG) of training vectors. Each vector is integrated at the PPC, giving rise to a movement command. The generation of each vector induces a complementary postural phase during which ERG output stops and learning occurs. Then a new vector is generated and the cycle is repeated. This cyclic, biphasic behavior is controlled by a specialized gated dipole circuit. ERG output autonomously stops in such a way that, across trials, a broad sample of workspace target positions is generated. When the ERG shuts off, a modulator gate opens, copying the PPC into the TPC. Learning of a transformation from TPC to PPC occurs using the DV as an error signal that is zeroed due to learning. This learning scheme is called a Vector Associative Map, or VAM. The VAM model is a general-purpose device for autonomous real-time error-based learning and performance of associative maps. The DV stage serves the dual function of reading out new TPCs during performance and reading in new adaptive weights during learning, without a disruption of real-time operation. YAMs thus provide an on-line unsupervised alternative to the off-line properties of supervised error-correction learning algorithms. YAMs and VAM cascades for learning motor-to-motor and spatial-to-motor maps are described. YAM models and Adaptive Resonance Theory (ART) models exhibit complementary matching, learning, and performance properties that together provide a foundation for designing a total sensory-cognitive and cognitive-motor autonomous system.

Altered ultrasonic vocalization and impaired learning and memory in Angelman syndrome mouse model with a large maternal deletion from Ube3a to Gabrb3.

Angelman syndrome (AS) is a neurobehavioral disorder associated with mental retardation, absence of language development, characteristic electroencephalography (EEG) abnormalities and epilepsy, happy disposition, movement or balance disorders, and autistic behaviors. The molecular defects underlying AS are heterogeneous, including large maternal deletions of chromosome 15q11-q13 (70%), paternal uniparental disomy (UPD) of chromosome 15 (5%), imprinting mutations (rare), and mutations in the E6-AP ubiquitin ligase gene UBE3A (15%). Although patients with UBE3A mutations have a wide spectrum of neurological phenotypes, their features are usually milder than AS patients with deletions of 15q11-q13. Using a chromosomal engineering strategy, we generated mutant mice with a 1.6-Mb chromosomal deletion from Ube3a to Gabrb3, which inactivated the Ube3a and Gabrb3 genes and deleted the Atp10a gene. Homozygous deletion mutant mice died in the perinatal period due to a cleft palate resulting from the null mutation in Gabrb3 gene. Mice with a maternal deletion (m-/p+) were viable and did not have any obvious developmental defects. Expression analysis of the maternal and paternal deletion mice confirmed that the Ube3a gene is maternally expressed in brain, and showed that the Atp10a and Gabrb3 genes are biallelically expressed in all brain sub-regions studied. Maternal (m-/p+), but not paternal (m+/p-), deletion mice had increased spontaneous seizure activity and abnormal EEG. Extensive behavioral analyses revealed significant impairment in motor function, learning and memory tasks, and anxiety-related measures assayed in the light-dark box in maternal deletion but not paternal deletion mice. Ultrasonic vocalization (USV) recording in newborns revealed that maternal deletion pups emitted significantly more USVs than wild-type littermates. The increased USV in maternal deletion mice suggests abnormal signaling behavior between mothers and pups that may reflect abnormal communication behaviors in human AS patients. Thus, mutant mice with a maternal deletion from Ube3a to Gabrb3 provide an AS mouse model that is molecularly more similar to the contiguous gene deletion form of AS in humans than mice with Ube3a mutation alone. These mice will be valuable for future comparative studies to mice with maternal deficiency of Ube3a alone.

Of mice, birds, and men: the mouse ultrasonic song system has some features similar to humans and song-learning birds.

Humans and song-learning birds communicate acoustically using learned vocalizations. The characteristic features of this social communication behavior include vocal control by forebrain motor areas, a direct cortical projection to brainstem vocal motor neurons, and dependence on auditory feedback to develop and maintain learned vocalizations. These features have so far not been found in closely related primate and avian species that do not learn vocalizations. Male mice produce courtship ultrasonic vocalizations with acoustic features similar to songs of song-learning birds. However, it is assumed that mice lack a forebrain system for vocal modification and that their ultrasonic vocalizations are innate. Here we investigated the mouse song system and discovered that it includes a motor cortex region active during singing, that projects directly to brainstem vocal motor neurons and is necessary for keeping song more stereotyped and on pitch. We also discovered that male mice depend on auditory feedback to maintain some ultrasonic song features, and that sub-strains with differences in their songs can match each other's pitch when cross-housed under competitive social conditions. We conclude that male mice have some limited vocal modification abilities with at least some neuroanatomical features thought to be unique to humans and song-learning birds. To explain our findings, we propose a continuum hypothesis of vocal learning.

Advances in understanding mechanisms of thalamic relays in cognition and behavior.

The main impetus for a mini-symposium on corticothalamic interrelationships was the recent number of studies highlighting the role of the thalamus in aspects of cognition beyond sensory processing. The thalamus contributes to a range of basic cognitive behaviors that include learning and memory, inhibitory control, decision-making, and the control of visual orienting responses. Its functions are deeply intertwined with those of the better studied cortex, although the principles governing its coordination with the cortex remain opaque, particularly in higher-level aspects of cognition. How should the thalamus be viewed in the context of the rest of the brain? Although its role extends well beyond relaying of sensory information from the periphery, the main function of many of its subdivisions does appear to be that of a relay station, transmitting neural signals primarily to the cerebral cortex from a number of brain areas. In cognition, its main contribution may thus be to coordinate signals between diverse regions of the telencephalon, including the neocortex, hippocampus, amygdala, and striatum. This central coordination is further subject to considerable extrinsic control, for example, inhibition from the basal ganglia, zona incerta, and pretectal regions, and chemical modulation from ascending neurotransmitter systems. What follows is a brief review on the role of the thalamus in aspects of cognition and behavior, focusing on a summary of the topics covered in a mini-symposium held at the Society for Neuroscience meeting, 2014.

Assessing segmentation processes by click detection: online measure of statistical learning, or simple interference?

Statistical learning can be used to extract the words from continuous speech. Gómez, Bion, and Mehler (Language and Cognitive Processes, 26, 212–223, 2011) proposed an online measure of statistical learning: They superimposed auditory clicks on a continuous artificial speech stream made up of a random succession of trisyllabic nonwords. Participants were instructed to detect these clicks, which could be located either within or between words. The results showed that, over the length of exposure, reaction times (RTs) increased more for within-word than for between-word clicks. This result has been accounted for by means of statistical learning of the between-word boundaries. However, even though statistical learning occurs without an intention to learn, it nevertheless requires attentional resources. Therefore, this process could be affected by a concurrent task such as click detection. In the present study, we evaluated the extent to which the click detection task indeed reflects successful statistical learning. Our results suggest that the emergence of RT differences between within- and between-word click detection is neither systematic nor related to the successful segmentation of the artificial language. Therefore, instead of being an online measure of learning, the click detection task seems to interfere with the extraction of statistical regularities.

De la cooperativa agroalimentaria a la "learning netchain". Hacia un planteamiento teorico interorganizativo e interpersonal

Se propone un planteamiento teórico/conceptual para determinar si las relaciones interorganizativas e interpersonales de la netchain de las cooperativas agroalimentarias evolucionan hacia una learning netchain. Las propuestas del trabajo muestran que el mayor grado de asociacionismo y la mayor cooperación/colaboración vertical a lo largo de la cadena están positivamente relacionados con la posición horizontal de la empresa focal más cercana del consumidor final. Esto requiere una planificación y una resolución de problemas de manera conjunta, lo que está positivamente relacionado con el mayor flujo y diversidad de la información/conocimiento obtenido y diseminado a lo largo de la netchain. Al mismo tiempo se necesita desarrollar un contexto social en el que fluya la información/conocimiento y las nuevas ideas de manera informal y esto se logra con redes personales y, principalmente, profesionales y con redes internas y, principalmente, externas. Todo esto permitirá una mayor satisfacción de los socios de la cooperativa agroalimentaria y de sus distribuidores y una mayor intensidad en I+D, convirtiéndose la netchain de la cooperativa agroalimentaria, así, en una learning netchain.