Название: Artificial Intelligence Driven Circuits and Systems: Select Proceedings of ISED 2021
Автор: Biswajit Mishra, Jimson Mathew, Priyadarsan Patra
Издательство: Springer
Год: 2022
Страниц: 167
Язык: английский
Формат: pdf (true)
Размер: 10.2 MB

This book highlights selected papers presented at the 10th International Symposium on Embedded Computing and System Design (ISED) 2021. This symposium provides a platform for researchers to share the latest scientific achievements of embedded computing and system design. The book is divided into three broad sections. The first section discusses topics like VLSI and testing, circuits and systems with a focus on emerging technologies. The second section discusses topics like embedded hardware and software systems and novel applications. The final section discusses the state-of-the-art technologies involving IoT, artificial intelligence, green and edge computing that demonstrates the issues currently.

Source code documentation is the process of writing concise, natural language descriptions of how the source code behaves during run time. In the work "NeuralDoc-Automating Code Translation Using Machine Learning", we propose a novel approach called NeuralDoc, for automating source code documentation using Machine Learning techniques. We model automatic code documentation as a language translation task, where the source code serves as the input sequence, which is translated by the Machine Learning model to natural language sentences depicting the functionality of the program. The Machine Learning model that we use is the Transformer, which leverages the self-attention and multi-headed attention features to effectively capture long-range dependencies and has been shown to perform well on a range of natural language processing (NLP) tasks. We integrate the copy attention mechanism and incorporate the use of BERT, which is a pre-training technique into the basic Transformer architecture to create a novel approach for automating code documentation. We build an intuitive interface for users to interact with our models and deploy our system as a web application. We carry out experiments on two datasets consisting of Java and Python source programs and their documentation, to demonstrate the effectiveness of our proposed method.

The work "Transfer Fault Detection in Finite State Machines Using Deep Neural Networks" presents a testing scheme for Finite State Machines (FSM) based on Deep Neural Network (DNN). This technique determines whether a given implementation FSM-B is equivalent to its specification FSM-A. The input/output sequences (I/O pairs) for a given FSM train the proposed DNN. First, I/O pairs of FSM-A are generated using an adaptive distinguishing algorithm, and most of these sequences (around 80%) are used for training the DNN. After training, the remaining 20prc I/O pairs are used for validating the derived DNN. After training and validation, the correctness of FSM-B is checked. A small number of vectors is applied to FSM-B and the generated outputs are compared with the DNN predicted outputs. FSMs lie at the heart of various complex computing systems like sequential circuits, telecommunications systems, communications protocols, embedded systems and many other related fields. The control portions of today’s communication protocols are mostly modelled by FSM. These complex systems are also less reliable. Hence, testing such FSM is indispensable to ensure their correct functioning. An FSM has a finite set of states and produces output on state transitions upon receiving the input. In testing problems, an FSM consists of a transition diagram, however, its present state is not known. An input sequence known as homing sequence is applied to the FSM so that from its input–output behaviour, its state information can be derived.