Specialized Mappings Architecture with Applications to Vision-based Estimation of Articulated Body Pose

A fundamental task of vision systems is to infer the state of the world given some form of visual observations. From a computational perspective, this often involves facing an ill-posed problem; e.g., information is lost via projection of the 3D world into a 2D image. Solution of an ill-posed problem requires additional information, usually provided as a model of the underlying process. It is important that the model be both computationally feasible as well as theoretically well-founded. In this thesis, a probabilistic, nonlinear supervised computational learning model is proposed: the Specialized Mappings Architecture (SMA). The SMA framework is demonstrated in a computer vision system that can estimate the articulated pose parameters of a human body or human hands, given images obtained via one or more uncalibrated cameras. The SMA consists of several specialized forward mapping functions that are estimated automatically from training data, and a possibly known feedback function. Each specialized function maps certain domains of the input space (e.g., image features) onto the output space (e.g., articulated body parameters). A probabilistic model for the architecture is first formalized. Solutions to key algorithmic problems are then derived: simultaneous learning of the specialized domains along with the mapping functions, as well as performing inference given inputs and a feedback function. The SMA employs a variant of the Expectation-Maximization algorithm and approximate inference. The approach allows the use of alternative conditional independence assumptions for learning and inference, which are derived from a forward model and a feedback model. Experimental validation of the proposed approach is conducted in the task of estimating articulated body pose from image silhouettes. Accuracy and stability of the SMA framework is tested using artificial data sets, as well as synthetic and real video sequences of human bodies and hands.

Biased ART: A Neural Architecture that Shifts Attention Toward Previously Disregarded Features Following an Incorrect Prediction

Memories in Adaptive Resonance Theory (ART) networks are based on matched patterns that focus attention on those portions of bottom-up inputs that match active top-down expectations. While this learning strategy has proved successful for both brain models and applications, computational examples show that attention to early critical features may later distort memory representations during online fast learning. For supervised learning, biased ARTMAP (bARTMAP) solves the problem of over-emphasis on early critical features by directing attention away from previously attended features after the system makes a predictive error. Small-scale, hand-computed analog and binary examples illustrate key model dynamics. Twodimensional simulation examples demonstrate the evolution of bARTMAP memories as they are learned online. Benchmark simulations show that featural biasing also improves performance on large-scale examples. One example, which predicts movie genres and is based, in part, on the Netflix Prize database, was developed for this project. Both first principles and consistent performance improvements on all simulation studies suggest that featural biasing should be incorporated by default in all ARTMAP systems. Benchmark datasets and bARTMAP code are available from the CNS Technology Lab Website: http://techlab.bu.edu/bART/.

A Neural Network Architecture for Autonomous Learning, Recognition, and Prediction in a Nonstationary World

In a constantly changing world, humans are adapted to alternate routinely between attending to familiar objects and testing hypotheses about novel ones. We can rapidly learn to recognize and narne novel objects without unselectively disrupting our memories of familiar ones. We can notice fine details that differentiate nearly identical objects and generalize across broad classes of dissimilar objects. This chapter describes a class of self-organizing neural network architectures--called ARTMAP-- that are capable of fast, yet stable, on-line recognition learning, hypothesis testing, and naming in response to an arbitrary stream of input patterns (Carpenter, Grossberg, Markuzon, Reynolds, and Rosen, 1992; Carpenter, Grossberg, and Reynolds, 1991). The intrinsic stability of ARTMAP allows the system to learn incrementally for an unlimited period of time. System stability properties can be traced to the structure of its learned memories, which encode clusters of attended features into its recognition categories, rather than slow averages of category inputs. The level of detail in the learned attentional focus is determined moment-by-moment, depending on predictive success: an error due to over-generalization automatically focuses attention on additional input details enough of which are learned in a new recognition category so that the predictive error will not be repeated. An ARTMAP system creates an evolving map between a variable number of learned categories that compress one feature space (e.g., visual features) to learned categories of another feature space (e.g., auditory features). Input vectors can be either binary or analog. Computational properties of the networks enable them to perform significantly better in benchmark studies than alternative machine learning, genetic algorithm, or neural network models. Some of the critical problems that challenge and constrain any such autonomous learning system will next be illustrated. Design principles that work together to solve these problems are then outlined. These principles are realized in the ARTMAP architecture, which is specified as an algorithm. Finally, ARTMAP dynamics are illustrated by means of a series of benchmark simulations.

The Specialized Mappings Architecture

A probabilistic, nonlinear supervised learning model is proposed: the Specialized Mappings Architecture (SMA). The SMA employs a set of several forward mapping functions that are estimated automatically from training data. Each specialized function maps certain domains of the input space (e.g., image features) onto the output space (e.g., articulated body parameters). The SMA can model ambiguous, one-to-many mappings that may yield multiple valid output hypotheses. Once learned, the mapping functions generate a set of output hypotheses for a given input via a statistical inference procedure. The SMA inference procedure incorporates an inverse mapping or feedback function in evaluating the likelihood of each of the hypothesis. Possible feedback functions include computer graphics rendering routines that can generate images for given hypotheses. The SMA employs a variant of the Expectation-Maximization algorithm for simultaneous learning of the specialized domains along with the mapping functions, and approximate strategies for inference. The framework is demonstrated in a computer vision system that can estimate the articulated pose parameters of a human’s body or hands, given silhouettes from a single image. The accuracy and stability of the SMA are also tested using synthetic images of human bodies and hands, where ground truth is known.

Efficient End-Host Architecture for High Performance Communication Using User-level Sandboxing

Current low-level networking abstractions on modern operating systems are commonly implemented in the kernel to provide sufficient performance for general purpose applications. However, it is desirable for high performance applications to have more control over the networking subsystem to support optimizations for their specific needs. One approach is to allow networking services to be implemented at user-level. Unfortunately, this typically incurs costs due to scheduling overheads and unnecessary data copying via the kernel. In this paper, we describe a method to implement efficient application-specific network service extensions at user-level, that removes the cost of scheduling and provides protected access to lower-level system abstractions. We present a networking implementation that, with minor modifications to the Linux kernel, passes data between "sandboxed" extensions and the Ethernet device without copying or processing in the kernel. Using this mechanism, we put a customizable networking stack into a user-level sandbox and show how it can be used to efficiently process and forward data via proxies, or intermediate hosts, in the communication path of high performance data streams. Unlike other user-level networking implementations, our method makes no special hardware requirements to avoid unnecessary data copies. Results show that we achieve a substantial increase in throughput over comparable user-space methods using our networking stack implementation.

A Short History of Computational Complexity

National Science Foundation (CCR-998310); Army Research Office (DAAD19-02-1-0058)

Assessing the Security of a Clean-Slate Internet Architecture

The TCP/IP architecture was originally designed without taking security measures into consideration. Over the years, it has been subjected to many attacks, which has led to many patches to counter them. Our investigations into the fundamental principles of networking have shown that carefully following an abstract model of Interprocess Communication (IPC) addresses many problems [1]. Guided by this IPC principle, we designed a clean-slate Recursive INternet Architecture (RINA) [2]. In this paper, we show how, without the aid of cryptographic techniques, the bare-bones architecture of RINA can resist most of the security attacks faced by TCP/IP. We also show how hard it is for an intruder to compromise RINA. Then, we show how RINA inherently supports security policies in a more manageable, on-demand basis, in contrast to the rigid, piecemeal approach of TCP/IP.

Estimating Human Body Pose from a Single Image via the Specialized Mappings Architecture

A non-linear supervised learning architecture, the Specialized Mapping Architecture (SMA) and its application to articulated body pose reconstruction from single monocular images is described. The architecture is formed by a number of specialized mapping functions, each of them with the purpose of mapping certain portions (connected or not) of the input space, and a feedback matching process. A probabilistic model for the architecture is described along with a mechanism for learning its parameters. The learning problem is approached using a maximum likelihood estimation framework; we present Expectation Maximization (EM) algorithms for two different instances of the likelihood probability. Performance is characterized by estimating human body postures from low level visual features, showing promising results.

The Cyclone Server Architecture: Streamlining Delivery of Popular Content

We propose a new technique for efficiently delivering popular content from information repositories with bounded file caches. Our strategy relies on the use of fast erasure codes (a.k.a. forward error correcting codes) to generate encodings of popular files, of which only a small sliding window is cached at any time instant, even to satisfy an unbounded number of asynchronous requests for the file. Our approach capitalizes on concurrency to maximize sharing of state across different request threads while minimizing cache memory utilization. Additional reduction in resource requirements arises from providing for a lightweight version of the network stack. In this paper, we describe the design and implementation of our Cyclone server as a Linux kernel subsystem.

Computational Properties of SNAFU

Sensor applications in Sensoria [1] are expressed using STEP (Sensorium Task Execution Plan). SNAFU (Sensor-Net Applications as Functional Units) serves as a high-level sensor-programming language, which is compiled into STEP. In SNAFU’s current form, its differences with STEP are relatively minor, as they are limited to shorthands and macros not available in STEP. We show that, however restrictive it may seem, SNAFU has in fact universal power; technically, it is a Turing-complete language, i.e., any Turing program can be written in SNAFU (though not always conveniently). Although STEP may be allowed to have universal power, as a low-level language not directly available to Sensorium users, SNAFU programmers may use this power for malicious purposes or inadvertently introduce errors with destructive consequences. In future developments of SNAFU, we plan to introduce restrictions and highlevel features with safety guards, such as those provided by a type system, which will make SNAFU programming safer.

Extending snBench to Support a Video-Based Intrusion Detection and Alerting System with a Centralized Hash Table

In this project we design and implement a centralized hashing table in the snBench sensor network environment. We discuss the feasibility of this approach and compare and contrast with the distributed hashing architecture, with particular discussion regarding the conditions under which a centralized architecture makes sense. There are numerous computational tasks that require persistence of data in a sensor network environment. To help motivate the need for data storage in snBench we demonstrate a practical application of the technology whereby a video camera can monitor a room to detect the presence of a person and send an alert to the appropriate authorities.

KInNeSS: A Modular Framework for Computational Neuroscience

Making use of very detailed neurophysiological, anatomical, and behavioral data to build biological-realistic computational models of animal behavior is often a difficult task. Until recently, many software packages have tried to resolve this mismatched granularity with different approaches. This paper presents KInNeSS, the KDE Integrated NeuroSimulation Software environment, as an alternative solution to bridge the gap between data and model behavior. This open source neural simulation software package provides an expandable framework incorporating features such as ease of use, scalabiltiy, an XML based schema, and multiple levels of granularity within a modern object oriented programming design. KInNeSS is best suited to simulate networks of hundreds to thousands of branched multu-compartmental neurons with biophysical properties such as membrane potential, voltage-gated and ligand-gated channels, the presence of gap junctions of ionic diffusion, neuromodulation channel gating, the mechanism for habituative or depressive synapses, axonal delays, and synaptic plasticity. KInNeSS outputs include compartment membrane voltage, spikes, local-field potentials, and current source densities, as well as visualization of the behavior of a simulated agent. An explanation of the modeling philosophy and plug-in development is also presented. Further developement of KInNeSS is ongoing with the ultimate goal of creating a modular framework that will help researchers across different disciplines to effecitively collaborate using a modern neural simulation platform.

KInNeSS: A Modular Framework for Computational Neuroscience

Making use of very detailed neurophysiological, anatomical, and behavioral data to build biologically-realistic computational models of animal behavior is often a difficult task. Until recently, many software packages have tried to resolve this mismatched granularity with different approaches. This paper presents KInNeSS, the KDE Integrated NeuroSimulation Software environment, as an alternative solution to bridge the gap between data and model behavior. This open source neural simulation software package provides an expandable framework incorporating features such as ease of use, scalability, an XML based schema, and multiple levels of granularity within a modern object oriented programming design. KInNeSS is best suited to simulate networks of hundreds to thousands of branched multi-compartmental neurons with biophysical properties such as membrane potential, voltage-gated and ligand-gated channels, the presence of gap junctions or ionic diffusion, neuromodulation channel gating, the mechanism for habituative or depressive synapses, axonal delays, and synaptic plasticity. KInNeSS outputs include compartment membrane voltage, spikes, local-field potentials, and current source densities, as well as visualization of the behavior of a simulated agent. An explanation of the modeling philosophy and plug-in development is also presented. Further development of KInNeSS is ongoing with the ultimate goal of creating a modular framework that will help researchers across different disciplines to effectively collaborate using a modern neural simulation platform.

An Active Pattern Recognition Architecture for Mobile Robots

An active, attentionally-modulated recognition architecture is proposed for object recognition and scene analysis. The proposed architecture forms part of navigation and trajectory planning modules for mobile robots. Key characteristics of the system include movement planning and execution based on environmental factors and internal goal definitions. Real-time implementation of the system is based on space-variant representation of the visual field, as well as an optimal visual processing scheme utilizing separate and parallel channels for the extraction of boundaries and stimulus qualities. A spatial and temporal grouping module (VWM) allows for scene scanning, multi-object segmentation, and featural/object priming. VWM is used to modulate a tn~ectory formation module capable of redirecting the focus of spatial attention. Finally, an object recognition module based on adaptive resonance theory is interfaced through VWM to the visual processing module. The system is capable of using information from different modalities to disambiguate sensory input.

Fusion Artmap: A Neural Network Architecture for Multi-Channel Data Fusion and Classification

Fusion ARTMAP is a self-organizing neural network architecture for multi-channel, or multi-sensor, data fusion. Single-channel Fusion ARTMAP is functionally equivalent to Fuzzy ART during unsupervised learning and to Fuzzy ARTMAP during supervised learning. The network has a symmetric organization such that each channel can be dynamically configured to serve as either a data input or a teaching input to the system. An ART module forms a compressed recognition code within each channel. These codes, in turn, become inputs to a single ART system that organizes the global recognition code. When a predictive error occurs, a process called paraellel match tracking simultaneously raises vigilances in multiple ART modules until reset is triggered in one of them. Parallel match tracking hereby resets only that portion of the recognition code with the poorest match, or minimum predictive confidence. This internally controlled selective reset process is a type of credit assignment that creates a parsimoniously connected learned network. Fusion ARTMAP's multi-channel coding is illustrated by simulations of the Quadruped Mammal database.