As deep learning (DL) opens the way to many technological innovations in a wild range of fields, more and more researchers and developers from diverse domains start to take advantage of DLs. In many circumstances, a d...
详细信息
ISBN:
(纸本)9781728115191
As deep learning (DL) opens the way to many technological innovations in a wild range of fields, more and more researchers and developers from diverse domains start to take advantage of DLs. In many circumstances, a developer leverages a DL framework and programs the training software in the form of source code (e.g., Python, Java). However, not all of the developers across domains are skilled at programming. It is highly desirable to provide a way so that a developer could focus on how to design and optimize their DL systems instead of spending too much time on programming. To simplify the programming process towards saving time and effort especially for beginners, we propose and implement Deepvisual, a visual programming tool for the design and development of DL systems. Deepvisual represents each layer of a neural network as a component. A user can drag-and-drop components to design and build a DL model, after which the training code is automatically generated. Moreover, Deepvisual supports to extract the neural network architecture on the given source code as input. We implement Deepvisual as a PyCharm plugin and demonstrate its usefulness on two typical use cases.
These last years, we have witnessed considerable improvements in machine learning and deep learning. Many advanced techniques are now based on deep neural networks. Although many software libraries are available, the ...
详细信息
ISBN:
(数字)9781510626928
ISBN:
(纸本)9781510626928
These last years, we have witnessed considerable improvements in machine learning and deep learning. Many advanced techniques are now based on deep neural networks. Although many software libraries are available, the development of deep neural networks requires a good level of mathematical knowledge and high programming skills. In this work, we present a visual tool to help simplify the programming of deep learning networks. The developed framework DeepViP is comprised of a node editor that provides users with a toolbox representing different types of neural layers. It allows the connection between the different blocks and the configuration of important hyperparameters of each layer. Thus, speeding-up experimentation with different architectures. Additionally, the developed solution offers users the possibility to generate a python script of the designed network that can be run using specific libraries such as keras or tensorflow.
Software development more and more focuses on mobile and connected IoT solutions. Whereas a variety of frameworks is available for professional developers, also the need comes up to deliver apps fast for personal use ...
详细信息
ISBN:
(纸本)9781728133959
Software development more and more focuses on mobile and connected IoT solutions. Whereas a variety of frameworks is available for professional developers, also the need comes up to deliver apps fast for personal use or rapid prototyping. Innovative ideas, in whatever context, must be realizable without having a vast amount of resources or deep domain-knowledge needed. In this work we present Pocket Code, a free open source mobile visual coding framework for Android and iOS with various hardware extensions, enabling everyone to create powerful apps directly on mobiles in short time. Especially the included visual coding-bricks for Arduino boards and Raspberry Pis also allow more sophisticated programs not only in an educational context to be developed with Pocket Code. In contrast to existing solutions, the presented app does not require any PC setting and is therefore making the development of apps mobile and broadening it to an even wider audience.
Deep learning is one of the fastest growing technologies in computer science with a plethora of applications. But this unprecedented growth has so far been limited to the consumption of deep learning experts. The prim...
详细信息
ISBN:
(纸本)9781450377164
Deep learning is one of the fastest growing technologies in computer science with a plethora of applications. But this unprecedented growth has so far been limited to the consumption of deep learning experts. The primary challenge being a steep learning curve for learning the programming libraries and the lack of intuitive systems enabling non-experts to consume deep learning. Towards this goal, we study the effectiveness of a "no-code" paradigm for designing deep learning models. Particularly, a visual drag-and-drop interface is found more efficient when compared with the traditional programming and alternative visual programming paradigms. We conduct user studies of different expertise levels to measure the entry level barrier and the developer load across different programming paradigms. We obtain a System Usability Scale (SUS) of 90 and a NASA Task Load index (TLX) score of 21 for the proposed visual programming compared to 68 and 52, respectively, for the traditional programming methods.
With the development of medical technology and automatic control, researchers have been working on autofocus for decades in order to reduce the errors caused by artificial factors. The core of auto-focus technology is...
详细信息
ISBN:
(数字)9781728162836
ISBN:
(纸本)9781728162843
With the development of medical technology and automatic control, researchers have been working on autofocus for decades in order to reduce the errors caused by artificial factors. The core of auto-focus technology is combining the image sharpness evaluation function and one-dimensional function optimization. When the function gets the highest value, the system is in focus position. Several classical sharpness evaluation curves and optimization algorithms are compared in this paper to analyse their advantages and disadvantages, and this paper proposes a microscope auto-focus method with image gray histogram average and hill-climbing algorithm. An experiment is conducted by using an NI, a stepping motor, a mobile load platform, a microscope, a shooting unit and a computer monitor. The control system is built with Labview visual programming software that can easily analyse and process signals. The experimental results show that the proposed microscope auto-focus method in this paper is better than those existing techniques by enhancing the accuracy and speed of auto-focus.
The paper introduces a visual programming language and corresponding web and cloud-based development environment called NetsBlox. NetsBlox is an extension of Snap! and builds upon its visual formalism as well as its o...
详细信息
The paper introduces a visual programming language and corresponding web and cloud-based development environment called NetsBlox. NetsBlox is an extension of Snap! and builds upon its visual formalism as well as its open source code base. NetsBlox adds distributed programming capabilities by introducing two well-known abstractions to block-based programming: message passing and Remote Procedure Calls (RPC). Messages containing data can be exchanged by two or more NetsBlox programs running on different computers connected to the Internet. RPCs are called on a client program and are executed on the NetsBlox server. These two abstractions make it possible to create distributed programs such as multi-player games or client-server applications. We believe that NetsBlox not only teaches basic distributed programming concepts but also provides increased motivation for high-school students to become creators and not just consumers of technology. (C) 2018 Elsevier Inc. All rights reserved.
Block-based visual programming (BVP) is becoming popular as a basis of programming education. It allows beginners to visually construct programs without suffering from syntax errors. However, a typical user interface ...
详细信息
ISBN:
(纸本)9781538642351
Block-based visual programming (BVP) is becoming popular as a basis of programming education. It allows beginners to visually construct programs without suffering from syntax errors. However, a typical user interface for BVP is inefficient partly because the users need to perform many drag-and-drop operations to put blocks on a program, and also partly because they need to find necessary blocks from many choices. To improve the efficiency of constructing programs in a BVP system, we propose a user interface that introduces three new features: (1) the semiautomatic addition of blocks;(2) the use of a pie menu to change categories of blocks;(3) the focus+context visualization of blocks in a category. We implemented a prototype BVP system with the new user interface.
This study explored the role of visual programming activities in computational thinking. A group of 158 freshmen majoring in information science and communication in Northern Taiwan were engaged in this study. A visua...
详细信息
ISBN:
(纸本)9781538650592
This study explored the role of visual programming activities in computational thinking. A group of 158 freshmen majoring in information science and communication in Northern Taiwan were engaged in this study. A visual programming platform, computational thinking questionnaire and a programming examination were employed to collect the log data of programming activities, self-responses to computational thinking and the score of programming examination for later analyses. The results reveal that visual programming activities would be critical to the evolution of multiple computational thinking perspectives.
A core challenge in introductory programming courses during the freshmen year is getting students to understand how a static textual representation (source code) maps to a highly dynamic process (program execution). M...
详细信息
ISBN:
(纸本)9781450351034
A core challenge in introductory programming courses during the freshmen year is getting students to understand how a static textual representation (source code) maps to a highly dynamic process (program execution). Most students nowadays are visual learners who learn programming concept better through web-based visual and interactive learning instead of learning from traditional black board lecturing styles. This presentation will report our experience with positive learning outcomes in teaching an introductory programming course in Python by using a dynamic visual programming development environment based on flow-charts (RAPTOR) and active learning with an interactive eBook (zyBook). In this study, eleven sections of COSC 111 were included over three semesters as control and experimental groups. Analysis showed a marginally significant difference (F (1,5) = 5.87; p = 0.060) between sections taught via the traditional method and sections taught with the proposed pedagogical innovation. We also found that there are statistically significant positive correlations in between uses of the interactive book (eBook) and performances of students in tests. To add to our understanding of what students were experiencing, we also administered a survey to students at the end of the course. Regarding learning styles and tools, survey result showed that the eBook was helpful in understanding programming concepts (71.5%), that the instant feedback that the online book provided was helpful (82.9%), that the class discussion and interaction were helpful (85.7%). Overall, it seems that the proposed pedagogical approaches have made a positive difference by increasing student motivation and engagement, and reducing failure rates.
Recently, model checkers, such as SPIN, have played an important role in the enhancement of software reliability. To promote the use of model checkers, we propose a visual programming language for SPIN model checkers ...
详细信息
ISBN:
(纸本)9783319699264;9783319699257
Recently, model checkers, such as SPIN, have played an important role in the enhancement of software reliability. To promote the use of model checkers, we propose a visual programming language for SPIN model checkers for educational use. Our prototype is based on Google Blockly.
暂无评论