Kaiserslautern - Fachbereich Informatik
Refine
Year of publication
- 1999 (267)
- 1996 (50)
- 1994 (49)
- 1995 (48)
- 1998 (38)
- 1997 (35)
- 2021 (27)
- 2016 (25)
- 2019 (24)
- 2022 (24)
- 1993 (22)
- 2015 (22)
- 2023 (22)
- 2001 (21)
- 2020 (21)
- 2007 (19)
- 2013 (18)
- 2018 (18)
- 2002 (17)
- 2003 (16)
- 2014 (15)
- 2012 (14)
- 1992 (13)
- 2000 (13)
- 2004 (12)
- 2006 (11)
- 2009 (11)
- 2008 (9)
- 2005 (8)
- 2017 (8)
- 1991 (7)
- 2010 (7)
- 2024 (7)
- 2011 (5)
- 1979 (2)
- 1980 (1)
- 1983 (1)
- 1990 (1)
Document Type
- Preprint (346)
- Doctoral Thesis (235)
- Report (139)
- Article (134)
- Master's Thesis (45)
- Study Thesis (13)
- Conference Proceeding (8)
- Bachelor Thesis (4)
- Part of a Book (2)
- Habilitation (2)
Has Fulltext
- yes (928)
Keywords
- AG-RESY (64)
- PARO (31)
- Case-Based Reasoning (20)
- Visualisierung (20)
- SKALP (16)
- CoMo-Kit (15)
- Fallbasiertes Schliessen (12)
- RODEO (12)
- Robotik (12)
- HANDFLEX (11)
Faculty / Organisational entity
Even though it is not very often admitted, partial functionsdo play a significant role in many practical applications of deduction sys-tems. Kleene has already given a semantic account of partial functionsusing a three-valued logic decades ago, but there has not been a satisfact-ory mechanization. Recent years have seen a thorough investigation ofthe framework of many-valued truth-functional logics. However, strongKleene logic, where quantification is restricted and therefore not truth-functional, does not fit the framework directly. We solve this problemby applying recent methods from sorted logics. This paper presents atableau calculus that combines the proper treatment of partial functionswith the efficiency of sorted calculi.
A Tailored Real Time Temporal Logic for Specifying Requirements of Building Automation Systems
(1999)
A tailored real time temporal logic for specifying requirements of building automation systems is introduced and analyzed. The logic features several new real time operators, which are chosen with regard to the application area. The new operators improve the conciseness and readability of requirements as compared to a general-purpose real time temporal logic. In addition, some of the operators also enhance the expressiveness of the logic. A number of properties of the new operators are presented and proven.
The rapid development of any field of knowledge brings with it unavoidable fragmentation and proliferation of new disciplines. The development of computer science is no exception. Software engineering (SE) and human-computer interaction (HCI) are both relatively new disciplines of computer science. Furthermore, as both names suggest, they each have strong connections with other subjects. SE is concerned with methods and tools for general software development based on engineering principles. This discipline has its roots not only in computer science but also in a number of traditional engineering disciplines. HCI is concerned with methods and tools for the development of human-computer interfaces, assessing the usability of computer systems and with broader issues about how people interact with computers. It is based on theories about how humans process information and interact with computers, other objects and other people in the organizational and social contexts in
which computers are used. HCI draws on knowledge and skills from psychology, anthropology and sociology in addition to computer science. Both disciplines need ways of measuring how well their products and development processes fulfil their intended requirements. Traditionally SE has been concerned with 'how software is constructed' and HCI with 'how people use software'. Given the
different histories of the disciplines and their different objectives, it is not surprising that they take different approaches to measurement. Thus, each has its own distinct 'measurement culture.' In this paper we analyse the differences and the commonalties of the two cultures by examining the measurement approaches used by each. We then argue the need for a common measurement taxonomy and framework, which is derived from our analyses of the two disciplines. Next we demonstrate the usefulness of the taxonomy and framework via specific example studies drawn from our own work and that of others and show that, in fact, the two disciplines have many important similarities as well as differences and that there is some evidence to suggest that they are growing closer. Finally, we discuss the role of the taxonomy as a framework to support: reuse, planning future studies, guiding practice and facilitating communication between the two disciplines.
In this work we propose a set of term-rewriting techniques for modelling object-oriented computation. Based on symbolic variants of explicit substitutions calculi, we show how to deal with imperative statements like assignment and sequence in specifications in a pure declarative style. Under our model, computation with classes and objects becomes simply normal form calculation, exactly as it is the case in term-rewriting based languages (for instance the functional languages). We believe this kind of unification between functions and
objects is important because it provides plausible alternatives for using the term-rewriting theory as an engine for supporting the formal and mechanical reasoning about object-oriented specifications.
Concept mapping is a simple and intuitive visual form of knowledge representation. Concept maps can be categorized as informal or formal, where the latter is characterized by implementing a semantics model constraining their components. Software engineering is a domain that has successfully adopted formal concept maps to visualize and specify complex systems. Automated tools have been implemented to support these models although their semantic constraints are hardcoded within the systems and hidden from users. This paper presents the Constraint Graphs and jKSImapper systems. Constraint Graphs is a flexible and powerful graphical system interface for specifying concept mapping notations. In addition, jKSImapper is a multi-user concept mapping editor for the Internet and the World Wide Web. Together, these systems aim to support user-definable formal concept mapping notations and distributed collaboration on the Internet and the World Wide Web.
The CBR team of the LISA is involved in several applied research projects based on the CBR paradigm. These applications use adaptation to solve the specific problems they face. So, we have capitalized some experience about how can be expressed and formalized adaptation processes. The bibliography on the subject is quite important but demonstrates a lake of formalism. At most, there exists some classifications about different types of adaptation.
We present a way to describe Reason Maintenance Systems using the sameformalism for justification based as well as for assumption based approaches.This formalism uses labelled formulae and thus is a special case of Gabbay'slabelled deductive systems. Since our approach is logic based, we are able toget a semantics oriented description of the systems in question.Instead of restricting ourselves to e.g. propositional Horn formulae, as wasdone in the past, we admit arbitrary logics. This enables us to characterizesystems as a whole, including both the reason maintenance component and theproblem solver, nevertheless maintaining a separation between the basic logicand the part that describes the label propagation. The possibility to freely varythe basic logic enables us to not only describe various existing systems, but canhelp in the design of completely new ones.We also show, that it is possible to implement systems based directly on ourlabelled logic and plead for "incremental calculi" crafted to attack undecidablelogics.Furthermore it is shown that the same approach can be used to handledefault reasoning, if the propositional labels are upgraded to first order.
The problem to interpolate Hermite-type data (i.e. two points with attached tangent vectors) with elastic curves of prescribed tension is known to have multiple solutions. A method is presented that finds all solutions of length not exceeding one period of its curvature function. The algorithm is based on algebraic relations between discrete curvature information which allow to transform the problem into a univariate one. The method operates with curves that by construction partially interpolate the given data. Hereby the objective function of the problem is drastically simplified. A bound on the maximum curvature value is established that provides an interval containing all solutions.
Caching has long been used to reduce average access latency, from registers and memory pages cached by hardware, to the application level such as a web browser retaining retrieved documents. We focus here on the high-level caching of potentially shared networked documents and define two terms in relation to this type of caching: Zero latency refers to the condition where access to a document produces a cache hit on the local machine, that is, there is little or no latency due to the network (we assume that latency due to local disk and memory access is insignificant in comparison to network latency). A document with zero latency usually has been placed in the cache after a previous access, or has been pulled there through some prefetching mechanism. Negative latency refers to automatic presentation, or push, of a document to a user based on a prediction that the user will want that document. With an ideal system, a user would be presented with documents either that she was about to request, or that she would not know to request but that would be immediately useful to her.
We will answer a question posed in [DJK91], and will show that Huet's completion algorithm [Hu81] becomes incomplete, i.e. it may generate a term rewriting system that is not confluent, if it is modified in a way that the reduction ordering used for completion can be changed during completion provided that the new ordering is compatible with the actual rules. In particular, we will show that this problem may not only arise if the modified completion algorithm does not terminate: Even if the algorithm terminates without failure, the generated finite noetherian term rewriting system may be non-confluent. Most existing implementations of the Knuth-Bendix algorithm provide the user with help in choosing a reduction ordering: If an unorientable equation is encountered, then the user has many options, especially, the one to orient the equation manually. The integration of this feature is based on the widespread assumption that, if equations are oriented by hand during completion and the completion process terminates with success, then the generated finite system is a maybe non terminating but locally confluent system (see e.g. [KZ89]). Our examples will show that this assumption is not true.
Computer-based simulation and visualization of acoustics of a virtual scene can aid during the design process of concert halls, lecture rooms, theaters, or living rooms. Because, not only the visual aspect of the room is important, but also its acoustics. In factory floors noise reduction is important since noise is hazardous to health. Despite the obvious dissimilarity between our aural and visual senses, many techniques required for the visualization of photo-realistic images and for the auralization of acoustic environments are quite similar. Both applications can be served by geometric methods such as particle- and ray tracing if we neglect a number of less important effects. By means of the simulation of room acoustics we want to predict the acoustic properties of a virtual model. For auralization, a pulse response filter needs to be assembled for each pair of source and listener positions. The convolution of this filter with an anechoic source signal provides the signal received at the listener position. Hence, the pulse response filter must contain all reverberations (echos) of a unit pulse, including their frequency decompositions due to absorption at different surface materials. For the room acoustic simulation a method named phonon tracing, since it is based on particles, is developed. The approach computes the energy or pressure decomposition for each particle (phonon) sent out from a sound source and uses this in a second pass (phonon collection) to construct the response filters for different listeners. This step can be performed in different precision levels. During the tracing step particle paths and additional information are stored in a so called phonon map. Using this map several sound visualization approaches were developed. From the visualization, the effect of different materials on the spectral energy / pressure distribution can be observed. The first few reflections already show whether certain frequency bands are rapidly absorbed. The absorbing materials can be identified and replaced in the virtual model, improving the overall acoustic quality of the simulated room. Furthermore an insight into the pressure / energy received at the listener position is possible. The phonon tracing algorithm as well as several sound visualization approaches are integrated into a common system utilizing Virtual Reality technologies in order to facilitate the immersion into the virtual scene. The system is a prototype developed within a project at the University of Kaiserslautern and is still a subject of further improvements. It consists of a stereoscopic back-projection system for visual rendering as well as professional audio equipment for auralization purposes.
Three dimensional (3d) point data is used in industry for measurement and reverse engineering. Precise point data is usually acquired with triangulating laser scanners or high precision structured light scanners. Lower precision point data is acquired by real-time structured light devices or by stereo matching with multiple cameras. The basic principle of all these methods is the so-called triangulation of 3d coordinates from two dimensional (2d) camera images.
This dissertation contributes a method for multi-camera stereo matching that uses a system of four synchronized cameras. A GPU based stereo matching method is presented to achieve a high quality reconstruction at interactive frame rates. Good depth resolution is achieved by allowing large disparities between the images. A multi level approach on the GPU allows a fast processing of these large disparities. In reverse engineering, hand-held laser scanners are used for the scanning of complex shaped objects. The operator of the scanner can scan complex regions slower, multiple times, or from multiple angles to achieve a higher point density. Traditionally, computer aided design (CAD) geometry is reconstructed in a separate step after the scanning. Errors or missing parts in the scan prevent a successful reconstruction. The contribution of this dissertation is an on-line algorithm that allows the reconstruction during the scanning of an object. Scanned points are added to the reconstruction and improve it on-line. The operator can detect the areas in the scan where the reconstruction needs additional data.
First, the point data is thinned out using an octree based data structure. Local normals and principal curvatures are estimated for the reduced set of points. These local geometric values are used for segmentation using a region growing approach. Implicit quadrics are fitted to these segments. The canonical form of the quadrics provides the parameters of basic geometric primitives.
An improved approach uses so called accumulated means of local geometric properties to perform segmentation and primitive reconstruction in a single step. Local geometric values can be added and removed on-line to these means to get a stable estimate over a complete segment. By estimating the shape of the segment it is decided which local areas are added to a segment. An accumulated score estimates the probability for a segment to belong to a certain type of geometric primitive. A boundary around the segment is reconstructed using a growing algorithm that ensures that the boundary is closed and avoids self intersections.
The reasoning power of human-oriented plan-based reasoning systems is primarilyderived from their domain-specific problem solving knowledge. Such knowledge is, how-ever, intrinsically incomplete. In order to model the human ability of adapting existingmethods to new situations we present in this work a declarative approach for represent-ing methods, which can be adapted by so-called meta-methods. Since apparently thesuccess of this approach relies on the existence of general and strong meta-methods,we describe several meta-methods of general interest in detail by presenting the prob-lem solving process of two familiar classes of mathematical problems. These examplesshould illustrate our philosophy of proof planning as well: besides planning with thecurrent repertoire of methods, the repertoire of methods evolves with experience inthat new ones are created by meta-methods which modify existing ones.
INRECA offers tools and methods for developing, validating, and maintaining classification, diagnosis and decision support systems. INRECA's basic technologies are inductive and case-based reasoning [9]. INRECA fully integrates [2] both techniques within one environment and uses the respective advantages of both technologies. Its object-oriented representation language CASUEL [10, 3] allows the definition of complex case structures, relations, similarity measures, as well as background knowledge to be used for adaptation. The objectoriented representation language makes INRECA a domain independent tool for its destined kind of tasks. When problems are solved via case-based reasoning, the primary kind of knowledge that is used during problem solving is the very specific knowledge contained in the cases. However, in many situations this specific knowledge by itself is not sufficient or appropriate to cope with all requirements of an application. Very often, background knowledge is available and/or necessary to better explore and interpret the available cases [1]. Such general knowledge may state dependencies between certain case features and can be used to infer additional, previously unknown features from the known ones.
In this paper we generalize the notion of method for proofplanning. While we adopt the general structure of methods introducedby Alan Bundy, we make an essential advancement in that we strictlyseparate the declarative knowledge from the procedural knowledge. Thischange of paradigm not only leads to representations easier to under-stand, it also enables modeling the important activity of formulatingmeta-methods, that is, operators that adapt the declarative part of exist-ing methods to suit novel situations. Thus this change of representationleads to a considerably strengthened planning mechanism.After presenting our declarative approach towards methods we describethe basic proof planning process with these. Then we define the notion ofmeta-method, provide an overview of practical examples and illustratehow meta-methods can be integrated into the planning process.
Extending the planADbased paradigm for auto-mated theorem proving, we developed in previ-ous work a declarative approach towards rep-resenting methods in a proof planning frame-work to support their mechanical modification.This paper presents a detailed study of a classof particular methods, embodying variations ofa mathematical technique called diagonaliza-tion. The purpose of this paper is mainly two-fold. First we demonstrate that typical math-ematical methods can be represented in ourframework in a natural way. Second we illus-trate our philosophy of proof planning: besidesplanning with a fixed repertoire of methods,metaADmethods create new methods by modify-ing existing ones. With the help of three differ-ent diagonalization problems we present an ex-ample trace protocol of the evolution of meth-ods: an initial method is extracted from a par-ticular successful proof. This initial method isthen reformulated for the subsequent problems,and more general methods can be obtained byabstracting existing methods. Finally we comeup with a fairly abstract method capable ofdealing with all the three problems, since it cap-tures the very key idea of diagonalization.
Adaptive Extraction and Representation of Geometric Structures from Unorganized 3D Point Sets
(2009)
The primary emphasis of this thesis concerns the extraction and representation of intrinsic properties of three-dimensional (3D) unorganized point clouds. The points establishing a point cloud as it mainly emerges from LiDaR (Light Detection and Ranging) scan devices or by reconstruction from two-dimensional (2D) image series represent discrete samples of real world objects. Depending on the type of scenery the data is generated from the resulting point cloud may exhibit a variety of different structures. Especially, in the case of environmental LiDaR scans the complexity of the corresponding point clouds is relatively high. Hence, finding new techniques allowing the efficient extraction and representation of the underlying structural entities becomes an important research issue of recent interest. This thesis introduces new methods regarding the extraction and visualization of structural features like surfaces and curves (e.g. ridge-lines, creases) from 3D (environmental) point clouds. One main part concerns the extraction of curve-like features from environmental point data sets. It provides a new method supporting a stable feature extraction by incorporating a probability-based point classification scheme that characterizes individual points regarding their affiliation to surface-, curve- and volume-like structures. Another part is concerned with the surface reconstruction from (environmental) point clouds exhibiting objects that are more or less complex. A new method providing multi-resolutional surface representations from regular point clouds is discussed. Following the applied principles of this approach a volumetric surface reconstruction method based on the proposed classification scheme is introduced. It allows the reconstruction of surfaces from highly unstructured and noisy point data sets. Furthermore, contributions in the field of reconstructing 3D point clouds from 2D image series are provided. In addition, a discussion concerning the most important properties of (environmental) point clouds with respect to feature extraction is presented.
This document introduces the extension of Katja to support position structures and explains the subtleties of their application as well as the design decisions made and problems solved with respect to their implementation. The Katja system was first introduced by Jan Schäfer in the context of his project work and is based on the MAX system developed by Arnd Poetzsch-Heffter.
If gradient based derivative algorithms are used to improve industrial products by reducing their target functions, the derivatives need to be exact.
The last percent of possible improvement, like the efficiency of a turbine, can only be gained if the derivatives are consistent with the solution process that is used in the simulation software.
It is problematic that the development of the simulation software is an ongoing process which leads to the use of approximated derivatives.
If a derivative computation is implemented manually, it will be inconsistent after some time if it is not updated.
This thesis presents a generalized approach which differentiates the whole simulation software with Algorithmic Differentiation (AD), and guarantees a correct and consistent derivative computation after each change to the software.
For this purpose, the variable tagging technique is developed.
The technique checks at run-time if all dependencies, which are used by the derivative algorithms, are correct.
Since it is also necessary to check the correctness of the implementation, a theorem is developed which describes how AD derivatives can be compared.
This theorem is used to develop further methods that can detect and correct errors.
All methods are designed such that they can be applied in real world applications and are used within industrial configurations.
The process described above yields consistent and correct derivatives but the efficiency can still be improved.
This is done by deriving new derivative algorithms.
A fixed-point iterator approach, with a consistent derivation, yields all state of the art algorithms and produces two new algorithms.
These two new algorithms include all implementation details and therefore they produce consistent derivative results.
For detecting hot spots in the application, the state of the art techniques are presented and extended.
The data management is changed such that the performance of the software is affected only marginally when quantities, like the number of input and output variables or the memory consumption, are computed for the detection.
The hot spots can be treated with techniques like checkpointing or preaccumulation.
How these techniques change the time and memory consumption is analyzed and it is shown how they need to be used in selected AD tools.
As a last step, the used AD tools are analyzed in more detail.
The major implementation strategies for operator overloading AD tools are presented and implementation improvements for existing AD tools are discussed.
The discussion focuses on a minimal memory consumption and makes it possible to compare AD tools on a theoretical level.
The new AD tool CoDiPack is based on these findings and its design and concepts are presented.
The improvements and findings in this thesis make it possible, that an automatic, consistent and correct derivative is generated in an efficient way for industrial applications.
Automated theorem proving is a search problem and, by its undecidability, a very difficult one. The challenge in the development of a practically successful prover is the mapping of the extensively developed theory into a program that runs efficiently on a computer. Starting from a level-based system model for automated theorem provers, in this work we present different techniques that are important for the development of powerful equational theorem provers. The contributions can be divided into three areas: Architecture. We present a novel prover architecture that is based on a set-based compression scheme. With moderate additional computational costs we achieve a substantial reduction of the memory requirements. Further wins are architectural clarity, the easy provision of proof objects, and a new way to parallelize a prover which shows respectable speed-ups in practice. The compact representation paves the way to new applications of automated equational provers in the area of verification systems. Algorithms. To improve the speed of a prover we need efficient solutions for the most time-consuming sub-tasks. We demonstrate improvements of several orders of magnitude for two of the most widely used term orderings, LPO and KBO. Other important contributions are a novel generic unsatisfiability test for ordering constraints and, based on that, a sufficient ground reducibility criterion with an excellent cost-benefit ratio. Redundancy avoidance. The notion of redundancy is of central importance to justify simplifying inferences which are used to prune the search space. In our experience with unfailing completion, the usual notion of redundancy is not strong enough. In the presence of associativity and commutativity, the provers often get stuck enumerating equations that are permutations of each other. By extending and refining the proof ordering, many more equations can be shown redundant. Furthermore, our refinement of the unfailing completion approach allows us to use redundant equations for simplification without the need to consider them for generating inferences. We describe the efficient implementation of several redundancy criteria and experimentally investigate their influence on the proof search. The combination of these techniques results in a considerable improvement of the practical performance of a prover, which we demonstrate with extensive experiments for the automated theorem prover Waldmeister. The progress achieved allows the prover to solve problems that were previously out of reach. This considerably enhances the potential of the prover and opens up the way for new applications.
Stochastic Network Calculus (SNC) emerged from two branches in the late 90s:
the theory of effective bandwidths and its predecessor the Deterministic Network
Calculus (DNC). As such SNC’s goal is to analyze queueing networks and support
their design and control.
In contrast to queueing theory, which strives for similar goals, SNC uses in-
equalities to circumvent complex situations, such as stochastic dependencies or
non-Poisson arrivals. Leaving the objective to compute exact distributions behind,
SNC derives stochastic performance bounds. Such a bound would, for example,
guarantee a system’s maximal queue length that is violated by a known small prob-
ability only.
This work includes several contributions towards the theory of SNC. They are
sorted into four main contributions:
(1) The first chapters give a self-contained introduction to deterministic net-
work calculus and its two branches of stochastic extensions. The focus lies on the
notion of network operations. They allow to derive the performance bounds and
simplifying complex scenarios.
(2) The author created the first open-source tool to automate the steps of cal-
culating and optimizing MGF-based performance bounds. The tool automatically
calculates end-to-end performance bounds, via a symbolic approach. In a second
step, this solution is numerically optimized. A modular design allows the user to
implement their own functions, like traffic models or analysis methods.
(3) The problem of the initial modeling step is addressed with the development
of a statistical network calculus. In many applications the properties of included
elements are mostly unknown. To that end, assumptions about the underlying
processes are made and backed by measurement-based statistical methods. This
thesis presents a way to integrate possible modeling errors into the bounds of SNC.
As a byproduct a dynamic view on the system is obtained that allows SNC to adapt
to non-stationarities.
(4) Probabilistic bounds are fundamentally different from deterministic bounds:
While deterministic bounds hold for all times of the analyzed system, this is not
true for probabilistic bounds. Stochastic bounds, although still valid for every time
t, only hold for one time instance at once. Sample path bounds are only achieved by
using Boole’s inequality. This thesis presents an alternative method, by adapting
the theory of extreme values.
(5) A long standing problem of SNC is the construction of stochastic bounds
for a window flow controller. The corresponding problem for DNC had been solved
over a decade ago, but remained an open problem for SNC. This thesis presents
two methods for a successful application of SNC to the window flow controller.
Accelerating the maturation process within the software engineering discipline may result in boosts of development productivity. One way to enable this acceleration is to develop tools and processes to mimic evolution of traditional engineering disciplines. Principles established in traditional engineering disciplines represent high-level guidance to constructing these tools and processes. This paper discusses two principles found in the traditional engineering disciplines and how these principles can apply to mature the software engineering discipline. The discussion is concretized through description of the Collaborative Management Environment, a software system under collaborative development among several national laboratories.
This paper shows an approach to profit from type information about planning objects in a partial-order planner. The approach turns out to combine representational and computational advantages. On the one hand, type hierarchies allow better structuring of domain specifications. On the other hand, operators contain type constraints which reduce the search space of the planner as they partially achieve the functionality of filter conditions.
In recent years, the Internet has become a major source of visual information exchange. Popular social platforms have reported an average of 80 million photo uploads a day. These images, are often accompanied with a user provided text one-liner, called an image caption. Deep Learning techniques have made significant advances towards automatic generation of factual image captions. However, captions generated by humans are much more than mere factual image descriptions. This work takes a step towards enhancing a machine's ability to generate image captions with human-like properties. We name this field as Affective Image Captioning, to differentiate it from the other areas of research focused on generating factual descriptions.
To deepen our understanding of human generated captions, we first perform a large-scale Crowd-Sourcing study on a subset of Yahoo Flickr Creative Commons 100 Million Dataset (YFCC100M). Three thousand random image-caption pairs were evaluated by native English speakers w.r.t different dimensions like focus, intent, emotion, meaning, and visibility. Our findings indicate three important underlying properties of human captions: subjectivity, sentiment, and variability. Based on these results, we develop Deep Learning models to address each of these dimensions.
To address the subjectivity dimension, we propose the Focus-Aspect-Value (FAV) model (along with a new task of aspect-detection) to structure the process of capturing subjectivity. We also introduce a novel dataset, aspects-DB, following this way of modeling. To implement the model, we propose a novel architecture called Tensor Fusion. Our experiments show that Tensor Fusion outperforms the state-of-the-art cross residual networks (XResNet) in aspect-detection.
Towards the sentiment dimension, we propose two models:Concept & Syntax Transition Network (CAST) and Show & Tell with Emotions (STEM). The CAST model uses a graphical structure to generate sentiment. The STEM model uses a neural network to inject adjectives into a neutral caption. Achieving a high score of 93% with human evaluation, these models were selected as the top-3 at the ACMMM Grand Challenge 2016.
To address the last dimension, variability, we take a generative approach called Generative Adversarial Networks (GAN) along with multimodal fusion. Our modified GAN, with two discriminators, is trained using Reinforcement Learning. We also show that it is possible to control the properties of the generated caption-variations with an external signal. Using sentiment as the external signal, we show that we can easily outperform state-of-the-art sentiment caption models.
Integrated project management means that design and planning are interleaved with plan execution, allowing both the design and plan to be changed as necessary. This requires that the right effects of change are propagated through the plan and design. When this is distributed among designers and planners, no one may have all of the information to perform such propagation and it is important to identify what effects should be propagated to whom when. We describe a set of dependencies among plan and design elements that allow such notification by a set of message-passing software agents. The result is to provide a novel level of computer support for complex projects.
The development of algorithmic differentiation (AD) tools focuses mostly on handling floating point types in the target language. Taping optimizations in these tools mostly focus on specific operations like matrix vector products. Aggregated types like std::complex are usually handled by specifying the AD type as a template argument. This approach provides exact results, but prevents the use of expression templates. If AD tools are extended and specialized such that aggregated types can be added to the expression framework, then this will result in reduced memory utilization and improve the timing for applications where aggregated types such as complex number or matrix vector operations are used. Such an integration requires a reformulation of the stored data per expression and a rework of the tape evaluation process. We will demonstrate the overheads on a synthetic benchmark and show the improvement when aggregated types are handled properly by the expression framework of the AD tool.
Based on the experiences from an autonomous mobile robot project called MOBOT-III, we found hard realtime-constraints for the operating- system-design. ALBATROSS is "A flexible multi-tasking and realtime network-operating-system-kernel". The focusin this article is on a communication-scheme fulfilling the previous demanded assurances. The centralchapters discuss the shared buffer management and the way to design the communication architecture.Some further aspects beside the strict realtime-requirements like the possibilities to control and watch a running system, are mentioned. ALBATROSS is actually implemented on a multi-processor VMEbus-system.
Based on experiences from an autonomous mobile robot project called MOBOT -III, we found hard realtime-constraints for the operating-system-design. ALBATROSS is "A flexible multi-tasking and realtime network-operatingsystem-kernel", not limited to mobile- robot-projects only, but which might be useful also wherever you have to guarantee a high reliability of a realtime-system. The focus in this article is on a communication-scheme fulfilling the demanded (hard realtime-) assurances although not implying time-delays or jitters on the critical informationchannels. The central chapters discuss a locking-free shared buffer management, without the need for interrupts and a way to arrange the communication architecture in order to produce minimal protocol-overhead and short cycle-times. Most of the remaining communication-capacity (if there is any) is used for redundant transfers, increasing the reliability of the whole system. ALBATROSS is actually implemented on a multi-processor VMEbus-system.
Ultraschall ist eines der am häufigsten genutzen, bildgebenden Verfahren in der Kardiologie. Dies ist durch die günstige Erzeugung, die Nicht-Invasivität und die Unschädlichkeit für die Patienten begründet. Nachteilig an den existierenden Geräten ist der Umstand, daß lediglich zwei-dimensionale Bilder generiert werden können. Zusätzlich können diese Bilder aufgrund anatomischer Gegebenheiten nicht aus einer wahlfreien Position akquiriert werden. Dies erschwert die Analyse der Daten und folglich die Diagnose. Mit dieser Arbeit wurden neue, algorithmische Aspekte des vier-dimensionalen, kardiologischen Ultraschalls ausgehend von der Akquisition der Rohdaten, deren Synchronisation und Rekonstruktion bis hin zur Visualisierung bearbeitet. In einem zusätzlichen Kapitel wurde eine neue Technik zur weiteren Aufwertung der Visualisierung, sowie zur visuellen Bearbeitung der Ultraschalldaten entwickelt. Durch die hier entwickelten Verfahren ist es möglich bestimmte Einschränkungen des kardiologischen Ultraschalls aufzuheben oder zumindest zu mildern. Hierunter zählen vor allem die Einschränkung auf zwei-dimensionale Schnittbilder, sowie die eingeschränkte Sichtwahl.
Software is becoming increasingly concurrent: parallelization, decentralization, and reactivity necessitate asynchronous programming in which processes communicate by posting messages/tasks to others’ message/task buffers. Asynchronous programming has been widely used to build fast servers and routers, embedded systems and sensor networks, and is the basis of Web programming using Javascript. Languages such as Erlang and Scala have adopted asynchronous programming as a fundamental concept with which highly scalable and highly reliable distributed systems are built.
Asynchronous programs are challenging to implement correctly: the loose coupling between asynchronously executed tasks makes the control and data dependencies difficult to follow. Even subtle design and programming mistakes on the programs have the capability to introduce erroneous or divergent behaviors. As asynchronous programs are typically written to provide a reliable, high-performance infrastructure, there is a critical need for analysis techniques to guarantee their correctness.
In this dissertation, I provide scalable verification and testing tools to make asyn- chronous programs more reliable. I show that the combination of counter abstraction and partial order reduction is an effective approach for the verification of asynchronous systems by presenting PROVKEEPER and KUAI, two scalable verifiers for two types of asynchronous systems. I also provide a theoretical result that proves a counter-abstraction based algorithm called expand-enlarge-check, is an asymptotically optimal algorithm for the coverability problem of branching vector addition systems as which many asynchronous programs can be modeled. In addition, I present BBS and LLSPLAT, two testing tools for asynchronous programs that efficiently uncover many subtle memory violation bugs.
ALICE
(1994)
Due to its performance, the field of deep learning has gained a lot of attention, with neural networks succeeding in areas like \( \textit{Computer Vision} \) (CV), \( \textit{Neural Language Processing} \) (NLP), and \( \textit{Reinforcement Learning} \) (RL). However, high accuracy comes at a computational cost as larger networks require longer training time and no longer fit onto a single GPU. To reduce training costs, researchers are looking into the dynamics of different optimizers, in order to find ways to make training more efficient. Resource requirements can be limited by reducing model size during training or designing more efficient models that improve accuracy without increasing network size.
This thesis combines eigenvalue computation and high-dimensional loss surface visualization to study different optimizers and deep neural network models. Eigenvectors of different eigenvalues are computed, and the loss landscape and optimizer trajectory are projected onto the plane spanned by those eigenvectors. A new parallelization method for the stochastic Lanczos method is introduced, resulting in faster computation and thus enabling high-resolution videos of the trajectory and second-order information during neural network training. Additionally, the thesis presents the loss landscape between two minima along with the eigenvalue density spectrum at intermediate points for the first time.
Secondly, this thesis presents a regularization method for \( \textit{Generative Adversarial Networks} \) (GANs) that uses second-order information. The gradient during training is modified by subtracting the eigenvector direction of the biggest eigenvalue, preventing the network from falling into the steepest minima and avoiding mode collapse. The thesis also shows the full eigenvalue density spectra of GANs during training.
Thirdly, this thesis introduces ProxSGD, a proximal algorithm for neural network training that guarantees convergence to a stationary point and unifies multiple popular optimizers. Proximal gradients are used to find a closed-form solution to the problem of training neural networks with smooth and non-smooth regularizations, resulting in better sparsity and more efficient optimization. Experiments show that ProxSGD can find sparser networks while reaching the same accuracy as popular optimizers.
Lastly, this thesis unifies sparsity and \( \textit{neural architecture search} \) (NAS) through the framework of group sparsity. Group sparsity is achieved through \( \ell_{2,1} \)-regularization during training, allowing for filter and operation pruning to reduce model size with minimal sacrifice in accuracy. By grouping multiple operations together, group sparsity can be used for NAS as well. This approach is shown to be more robust while still achieving competitive accuracies compared to state-of-the-art methods.
This paper presents a new kind of abstraction, which has been developed for the purpose of proofplanning. The basic idea of this paper is to abstract a given theorem and to find an abstractproof of it. Once an abstract proof has been found, this proof has to be refined to a real proofof the original theorem. We present a goal oriented abstraction for the purpose of equality proofplanning, which is parameterized by common parts of the left- and right-hand sides of the givenequality. Therefore, this abstraction technique provides an abstract equality problem which ismore adequate than those generated by the abstractions known so far. The presented abstractionalso supports the heuristic search process based on the difference reduction paradigm. We give aformal definition of the abstract space including the objects and their manipulation. Furthermore,we prove some properties in order to allow an efficient implementation of the presented abstraction.
There are two general approaches to providing for isochronous streams in the current Internet. The first approach is the resource reservation approach through protocols such as RSVP, or ATM technology. This provides bandwidth guarantees, however, it also requires significant upgrading of resources in the underlying network. The other common approach is adaptive rate control where the end-system has control of its rate according to feedback from the client population. This approach cannot guarantee timely delivery and raises some scaling questions, however a properly implemented scheme does improve quality and it requires no changes to the underlying IP network. Hence, there exists a dichotomy of requirements ; 1. To cater for reservation protocols or 'hooks' for future reservation components, and 2. To provide an architecture which provides an application controlled QoS scheme, which scales to the size of the current Internet in a best- effort architecture.
An Adaptive and Dynamic Simulation Framework for Incremental, Collaborative Classifier Fusion
(2016)
Abstract. To investigate incremental collaborative classifier fusion techniques, we have developed a comprehensive simulation framework. It is highly flexible and customizable, and can be adapted to various settings and scenarios. The toolbox is realized as an extension to the NetLogo multi-agent based simulation environment using its comprehensive Java- API. The toolbox has been integrated in two di↵erent environments, one for demonstration purposes and another, modeled on persons using re- alistic motion data from Zurich, who are communicating in an ad hoc fashion using mobile devices.
An agent-based approach to managing distributed, multi-platform software development projects
(1999)
This paper describes work undertaken within the context of the P3 (Project and Process Prompter) Project which aims to develop the Prompter tool, a 'decision-support tool to assist in the planning and managing of a software development project'. Prompter will have the ability to help software project managers to assimilate best practice and 'know how' in the field of software project management and incorporate expert critiquing to assist with solving the complex problems associated with software project management. This paper focuses on Prompters agent- based approach to tackling the problems of distributed, platform independent support.
A method for efficiently handling associativity and commutativity (AC) in implementations of (equational) theorem provers without incorporating AC as an underlying theory will be presented. The key of substantial efficiency gains resides in a more suitable representation of permutation-equations (such as f(x,f(y,z))=f(y,f(z,x)) for instance). By representing these permutation-equations through permutations in the mathematical sense (i.e. bijective func- tions :{1,..,n} {1,..,n}), and by applying adapted and specialized inference rules, we can cope more appropriately with the fact that permutation-equations are playing a particular role. Moreover, a number of restrictions concerning application and generation of permuta- tion-equations can be found that would not be possible in this extent when treating permu- tation-equations just like any other equation. Thus, further improvements in efficiency can be achieved.
In 1978, Klop demonstrated that a rewrite system constructed by adding the untyped lambda calculus, which has the Church-Rosser property, to a Church-Rosser first-order algebraic rewrite system may not be Church-Rosser. In contrast, Breazu-Tannen recently showed that argumenting any Church-Rosser first-order algebraic rewrite system with the simply-typed lambda calculus results in a Church-Rosser rewrite system. In addition, Breazu-Tannen and Gallier have shown that the second-order polymorphic lambda calculus can be added to such rewrite systems without compromising the Church-Rosser property (for terms which can be provably typed). There are other systems for which a Church-Rosser result would be desirable, among them being X^t+SP+FIX, the simply-typed lambda calculus extended with surjective pairing and fixed points. This paper will show that Klop's untyped counterexample can be lifted to a typed system to demonstrate that X^t+SP+FIX is not Church-Rosser.
We propose an approach to the problem of proof control for our new first-order inductive theorem prover QuodLibet that is characterized by a great deal of flexibility w.r.t. the forms of proof control the prover supports. The approach is based on so-called (proof) tactics, i.e. proof control routines written in a special proof control language named QML. QuodLibet provides a set of tactics (in addition to the elementary inference rules), which range from tactics for trivial simplification steps to tactics representing comprehensive inductive proof strategies. Moreover, QuodLibet allows new tactics that are written by the user in QML to be integrated into the system to dynamically extend its functionality.
We describe a technique to make application programs fault tolerant. This techADnique is based on the concept of checkpointing from an active program to one ormore passive backup copies which serve as an abstraction of stable memory. Ifthe primary copy fails, one of the backup copies takes over and resumes processADing service requests. After each failure a new backup copy is created in order torestore the replication degree of the service. All mechanisms necessary to achieveand maintain fault tolerance can be added automatically to the code of a nonADfaulttolerant server, thus making fault tolerance completely transparent for the applicaADtion programmer.
Open distributed systems are a class of distributed systems where (i) only partial information about the environment, in which they are running, is present, (ii) new resources may become available at runtime, and (iii) a subsystem may become aware of other subsystems after some interaction. Modeling and implementing such systems correctly is a complex task due to the openness and the dynamicity aspects. One way to ensure that the resulting systems behave correctly is to utilize formal verification.
Formal verification requires an adequate semantic model of the implementation, a specification of the desired behavior, and a reasoning technique. The actor model is a semantic model that captures the challenging aspects of open distributed systems by utilizing actors as universal primitives to represent system entities and allowing them to create new actors and to communicate by sending directed messages as reply to received messages. To enable compositional reasoning, where the reasoning task is reduced to independent verification of the system parts, semantic entities at a higher level of abstraction than actors are needed.
This thesis proposes an automaton model and combines sound reasoning techniques to compositionally verify implementations of open actor systems. Based on I/O automata, the model allows automata to be created dynamically and captures dynamic changes in communication patterns. Each automaton represents either an actor or a group of actors. The specification of the desired behavior is given constructively as an automaton. As the basis for compositionality, we formalize a component notion based on the static structure of the implementation instead of the dynamic entities (the actors) occurring in the system execution. The reasoning proceeds in two stages. The first stage establishes the connection between the automata representing single actors and their implementation description by means of weakest liberal preconditions. The second stage employs this result as the basis for verifying whether a component specification is satisfied. The verification is done by building a simulation relation from the automaton representing the implementation to the component's automaton. Finally, we validate the compositional verification approach through a number of examples by proving correctness of their actor implementations with respect to system specifications.
It has been observed that for understanding the biological function of certain RNA molecules, one has to study joint secondary structures of interacting pairs of RNA. In this thesis, a new approach for predicting the joint structure is proposed and implemented. For this, we introduce the class of m-dimensional context-free grammars --- an extension of stochastic context-free grammars to multiple dimensions --- and present an Earley-style semiring parser for this class. Additionally, we develop and thoroughly discuss an implementation variant of Earley parsers tailored to efficiently handle dense grammars, which embraces the grammars used for structure prediction. A currently proposed partitioning scheme for joint secondary structures is transferred into a two-dimensional context-free grammar, which in turn is used as a stochastic model for RNA-RNA interaction. This model is trained on actual data and then used for predicting most likely joint structures for given RNA molecules. While this technique has been widely used for secondary structure prediction of single molecules, RNA-RNA interaction was hardly approached this way in the past. Although our parser has O(n^3 m^3) time complexity and O(n^2 m^2) space complexity for two RNA molecules of sizes n and m, it remains practically applicable for typical sizes if enough memory is available. Experiments show that our parser is much more efficient for this application than classical Earley parsers. Moreover the predictions of joint structures are comparable in quality to current energy minimization approaches.
An Efficient Automated Machine Learning Framework for Genomics and Proteomics Sequence Analysis
(2023)
Genomics and Proteomics sequence analyses are the scientific studies of understanding the language of Deoxyribonucleic Acid (DNA), Ribonucleic Acid (RNA) and protein biomolecules with an objective of controlling the production of proteins and understanding their core functionalities. It helps to detect chronic diseases in early stages, root causes of clinical changes, key genetic targets for pharmaceutical development and optimization of therapeutics for various age groups. Most Genomics and Proteomics sequence analysis work is performed using typical wet lab experimental approaches that make use of different genetic diagnostic technologies. However, these approaches are costly, time consuming, skill and labor intensive. Hence, these approaches slow down the process of developing an efficient and economical sequence analysis landscape essential to demystify a variety of cellular processes and functioning of biomolecules in living organisms. To empower manual wet lab experiment driven research, many machine learning based approaches have been developed in recent years. However, these approaches cannot be used in practical environment due to their limited performance. Considering the sensitive and inherently demanding nature of Genomics and Proteomics sequence
analysis which can have very far-reaching as well as serious repercussions on account of misdiagnosis, the main
objective of this research is to develop an efficient automated computational framework for Genomics and Proteomics sequence analysis using the predictive and prescriptive analytical powers of Artificial Intelligence (AI) to significantly improve healthcare operations.
The proposed framework is comprised of 3 main components namely sequence encoding, feature engineering and
discrete or continuous value predictor. The sequence encoding module is equipped with a variety of existing and newly developed sequence encoding algorithms that are capable of generating a rich statistical representation of DNA, RNA and protein raw sequences. The feature engineering module has diverse types of feature selection and dimensionality reduction approaches which can be used to generate the most effective feature space. Furthermore, the discrete and/or continuous value predictor module of the proposed framework contains a wide range of existing machine learning and newly developed deep learning regressors and classifiers. To evaluate the integrity and generalizability of the proposed framework, we have performed a large-scale experimentation over diverse types of Genomics and Proteomics sequence analysis tasks (i.e., DNA, RNA and proteins).
In Genomics analysis, Epigenetic modification detection is one of the key component. It helps clinical researchers and practitioners to distinguish normal cellular activities from malfunctioned ones, which can lead to diverse genetic disorders such as metabolic disorders, cancers, etc. To support this analysis, the proposed framework is used to solve the problem of DNA and Histone modification prediction where it has achieved state-of-the-art performance on 27 publicly available benchmark datasets of 17 different species with best accuracy of 97%. RNA sequence analysis is another vital component of Genomics sequence analysis where the identification of different coding and non-coding RNAs as well as their subcellular localization patterns help to demystify the functions of diverse RNAs, root causes of clinical changes, develop precision medicine and optimize therapeutics. To support this analysis, the proposed framework is utilized for non-coding RNA classification and multi-compartment RNA subcellular localization prediction. Where it achieved state-of-the-art performance on 10 publicly available benchmark datasets of Homo sapiens and Mus Musculus species with best accuracy of 98%.
Proteomics sequence analysis is essential to demystify the virus pathogenesis, host immunity responses, the way
proteins affect or are affected by the cell processes, their structure and core functionalities. To support this analysis, the proposed framework is used for host protein-protein and virus-host protein-protein interaction prediction. It has achieved state-of-the-art performance on 2 publicly available protein protein interaction datasets of Homo Sapiens and Mus Musculus species with best accuracy of 96% and 7 viral host protein protein interaction datasets of multiple hosts and viruses with best accuracy of 94%. Considering the performance and practical significance of proposed framework, we believe proposed framework will help researchers in developing cutting-edge practical applications for diverse Genomic and Proteomic sequence analyses tasks (i.e., DNA, RNA and proteins).
Multidisciplinary optimizations (MDOs) encompass optimization problems that combine different disciplines into a single optimization with the aim of converging towards a design that simultaneously fulfills multiple criteria. For example, considering both fluid and structural disciplines to obtain a shape that is not only aerodynamically efficient, but also respects structural constraints. Combined with CAD-based parametrizations, the optimization produces an improved, manufacturable shape. For turbomachinery applications, this method has been successfully applied using gradient-free optimization methods such as genetic algorithms, surrogate modeling, and others. While such algorithms can be easily applied without access to the source code, the number of iterations to converge is dependent on the number of design parameters. This results in high computational costs and limited design spaces. A competitive alternative is offered by gradient-based optimization algorithms combined with adjoint methods, where the computational complexity of the gradient calculation is no longer dependent on the number of design parameters, but rather on the number of outputs. Such methods have been extensively used in single-disciplinary aerodynamic optimizations using adjoint fluid solvers and CAD parametrizations. However, CAD-based MDOs leveraging adjoint methods are just beginning to emerge.
This thesis contributes to this field of research by setting up a CAD-based adjoint MDO framework for turbomachinery design using both fluid and structural disciplines. To achieve this, the von Kármán Institute’s existing CAD-based optimization framework cado is augmented by the development of a FEM-based structural solver which has been differentiated using the algorithmic differentiation tool CoDiPack of TU Kaiserslautern. While most of the code could be differentiated in a black-box fashion, special treatment is required for the iterative linear and eigenvalue solvers to ensure accuracy and reduce memory consumption. As a result, the solver has the capability of computing both stress and vibration gradients at a cost independent on the number of design parameters. For the presented application case of a radial turbine optimization, the full gradient calculation has a computational cost of approximately 3.14 times the cost of a primal run and the peak memory usage of approximately 2.76 times that of a primal run.
The FEM code leverages object-oriented design such that the same base structure can be reused for different purposes with minimal re-differentiation. This is demonstrated by considering a composite material test case where the gradients could be easily calculated with respect to an extended design space that includes material properties. Additionally, the structural solver is reused within a CAD-based mesh deformation, which propagates the structural FEM mesh gradients through to the CAD parameters. This closes the link between the CAD shape and FEM mesh. Finally, the MDO framework is applied by optimizing the aerodynamic efficiency of a radial turbine while respecting structural constraints.