Interactive sketch generation

In this paper, we propose an interactive system for generating artistic sketches from images, based on the stylized multiresolution B-spline curve model and the livewire contour tracing paradigm. Our multiresolution B-spline stroke model allows interactive and continuous control of style and shape of the stroke at any level of details. Especially, we introduce a novel mathematical paradigm called the wavelet frame to provide essential properties for multiresolution stroke editing, such as feature point preservation, locality, time-efficiency, good approximation, etc. The livewire stroke map construction leads the user-guided stroke to automatically lock on to the target contour, allowing fast and accurate sketch drawing. We classify the target contours as outlines and interior flow, and develop two respective livewire techniques based on extended graph formulation and vector flow field. Experimental results show that the proposed system facilitates quick and easy generation of artistic sketches of various styles.     

Interactive manipulation of three dimensional binary scenes

A general framework for the interactive display and manipulation of the three-dimensional binary scenes is discussed. The procedures for obtaining the shaded graphics display, specification of a secondary object to be used in the manipulative operations of union, subtraction and intersection are described. The methods are illustrated using real medical data as well as regular geometric objects.     

Interactive rigid body manipulation with obstacle contacts

The interactive manipulation of rigid objects in virtual reality environments requires an object behaviour which is at least physically plausible to be useful for applications such as interactive assembly simulation and virtual training. Physically plausible behaviour implies that collisions between simulated solid objects are taken into account and that the motion of objects with obstacle contacts can be controlled without force feedback mechanisms in an intuitively correct manner. We present a real time framework which enables the simulation of interactively controlled solid objects with a dynamically changing set of contact constraints. In this paper all contact configurations are replaced by a canonical set of point contacts which is updated dynamically. The basic step to determine the contact forces and object motion consists in the solution of a non-linear complementarity problem (NCP), which results from the unilateral contact conditions together with an adequate discretization of the corresponding differential equations of motion. © 1998 John Wiley & Sons, Ltd.     

Structured generation and manipulation of graphical objects

Graphic languages in the form of subroutine packages can be both efficient and comfortable to use if the language features are structured very carefully. This is demonstrated by the clear picture structure, the simple but powerful operations on graphical objects, and the flexible identification mechanism realized in the graphical subroutine package LLGL.     

Centrifugal generation and manipulation of droplet emulsions

This work for the first time describes a centrifugal technique for the production and manipulation of highly monodisperse water droplets (CV of droplet diameter below 2%) immersed in a continuous flow of immiscible oil. Within a given working range, droplet volumes (5–22 nL) and their mutual spacing is governed by the channel geometry and the frequency of rotation. Different regimes of liquid–liquid flows are presented. We also demonstrate capabilities like droplet splitting and sedimentation as well as the production of two colored droplets, thus setting the stage for a novel centrifugal platform for multiphase flows.     

A formal application model for code and test generation

The considered model is used in manual development of application specifications and is based on the Letichevsky theory of basic protocols and a respective symbolic verification tools. Means to limit the behavioral characteristics of the model still matching the source requirements are discussed. If the model is verified successfully, the executable code of the application and the respective test code are generated from the model. The technique of using the developed model is described.     

Software monitoring through formal specification animation

This paper presents a formal specification-based software monitoring approach that can dynamically and continuously monitor the behaviors of a target system and explicitly recognize undesirable behaviors in the implementation with respect to its formal specification. The key idea of our approach is in building a monitoring module that connects a specification animator with a program debugger. The requirements information about expected dynamic behaviors of the target system are gathered from the formal specification animator, while the actual behaviors of concrete implementations of the target system are obtained through the program debugger. Based on the information obtained from both sides, the judgement on the conformance of the concrete implementation with respect to the formal specification is made timely while the target system is running. Furthermore, the proposed formal specification-based software monitoring technique does not embed any instrumentation codes to the target system nor does it annotate the target system with any formal specifications. It can detect implementation errors in a real-time manner, and help the developers and users of the system to react to the problems before critical failure occurs.     

Validation of formal specifications through transformation and animation

A significant impediment to the uptake of formal refinement-based methods among practitioners is the challenge of validating that the formal specifications of these methods capture the desired intents. Animation of specifications is widely recognized as an effective way of addressing such validation. However, animation tools are unable to directly execute (and thus animate) the typical uses of several of the specification constructs often found in ideal formal specifications. To address this problem, we have developed transformation heuristics that, starting with an ideal formal specification, guide its conversion into an animatable form. We show several of these heuristics and address the need to prove that the application of these transformations preserves the relevant behavior of the original specification. Portions of several case studies illustrate this approach.     

Code generation based on formal BURS theory and heuristic search

BURS theory provides a powerful mechanism to efficiently generate pattern matches in a given expression tree. BURS, which stands for bottom-up rewrite system, is based on term rewrite systems, to which costs are added. We formalise the underlying theory, and derive an algorithm that computes all pattern matches. This algorithm terminates if the term rewrite system is finite. We couple this algorithm with the well-known search algorithm A that carries out pattern selection. The search algorithm is directed by a cost heuristic that estimates the minimum cost of code that has yet to be generated. The advantage of using a search algorithm is that we need to compute only those costs that may be part of an optimal rewrite sequence (and not the costs of all possible rewrite sequences as in dynamic programming). A system that implements the algorithms presented in this work has been built. Received: 20 November 1995 / 26 June 1996     

Parser generation and grammar manipulation using prolog's infinite trees

The construction of parsers recognizing strings in a context-free language L(G) is usually done by generating a set of PROLOG clauses capable of parsing strings of L. Although this is a convenient way of generating parsers, it does not readily allow one to study the properties of G or to perform grammar transformations. This paper proposes a general PROLOG parser capable of analyzing the strings of L(G) when G is presented as an infinite tree. In particular, if the syntax of the grammar rules themselves is specified by a grammar, the parser can be used to recognize context-free grammar rules. Subsequently, actions can be attached to the parser so that it is possible to generate infinite trees for any grammar G′. Strings of L(G′) can then be analyzed uing the same general parser. The proposed approach allows the verification of conditions such as left recursion, satisfiability of LL(1) requirements, as well as grammar transformations such as ε elimination. The same analyzer is also capable of generating efficient parsers for specific machines.     

Taming distributed system complexity through formal patterns

Many distributed systems are real-time, safety-critical systems with strong qualitative and quantitative formal requirements. They often need to be reflective and adaptive, and may be probabilistic in their algorithms and/or their operating environments. All this makes these systems quite complex and therefore hard to design, build and verify. To tame such system complexity, this paper proposes formal patterns, that is, formally specified solutions to frequently occurring distributed system problems that are generic, executable, and come with strong formal guarantees. The semantics of such patterns as theory transformations in rewriting logic is explained; and a representative collection of useful patterns is presented to ground all the key concepts and show their effectiveness.     

基于变分自编码器的交互式旋律生成方法

现有的交互式神经音乐生成方法主要存在控制模式不灵活、数据标注困难以及模型难以优化等问题。针对这些问题,提出了一种基于变分自编码器(VAE)的无监督交互式旋律生成方法。通过为VAE引入显式的旋律轮廓条件推理学习,实现了对生成旋律局部与全局特征的灵活控制。实验表明,该方法易于优化且具有良好的旋律局部与全局特征的控制能力。通过对大量生成样本的分析,证明了模型从音乐数据中学习到了有用的音乐知识。     

Robotic grasping and manipulation through human visuomotor learning

A major goal of robotics research is to develop techniques that allow non-experts to teach robots dexterous skills. In this paper, we report our progress on the development of a framework which exploits human sensorimotor learning capability to address this aim. The idea is to place the human operator in the robot control loop where he/she can intuitively control the robot, and by practice, learn to perform the target task with the robot. Subsequently, by analyzing the robot control obtained by the human, it is possible to design a controller that allows the robot to autonomously perform the task. First, we introduce this framework with the ball-swapping task where a robot hand has to swap the position of the balls without dropping them, and present new analyses investigating the intrinsic dimension of the ball-swapping skill obtained through this framework. Then, we present new experiments toward obtaining an autonomous grasp controller on an anthropomorphic robot. In the experiments, the operator directly controls the (simulated) robot using visual feedback to achieve robust grasping with the robot. The data collected is then analyzed for inferring the grasping strategy discovered by the human operator. Finally, a method to generalize grasping actions using the collected data is presented, which allows the robot to autonomously generate grasping actions for different orientations of the target object.