최윤석 교수

As mobile hand-held devices including mobile phone are required to provide multi-media services more and more, it is necessary that the various hardware including high speed memory, high capacity data storage device, and high performance logic processor are integrated into the limited volume, which results in high density 3D SIP. In this work, power delivery network for 3D SIP integrated on silicon interposer will be discussed. The silicon interposer used in 3D SIP includes integrated decoupling

We propose a model that learns both the sequential and the structural features of code for source code summarization. We adopt the abstract syntax tree (AST) and graph convolution to model the structural information and the Transformer to model the sequential information. We convert code snippets into ASTs and apply graph convolution to obtain structurally-encoded node representations. Then, the sequences of the graphconvolutioned AST nodes are processed by the Transformer layers. Since structur

Neuromorphic systems are emerging as a promising alternative to revolutionize silicon-based computing devices. Two-dimensional materials are considered promising candidates for active materials due to their unique advantages. This review provides a comprehensive overview of neuromorphic computing based on 2D materials, encompassing the development from the synaptic devices based on 2D materials to the neuromorphic system demonstration. Besides, we further outline various applications of neuromor

We present a CMOS image sensor (CIS) with phase detection auto-focus (PDAF) in all pixels. The size of photodiode (PD) is 0.64μm by 1.28μm, the smallest ever reported and two PDs compose a single pixel. Inter PD isolation was fabricated by deep trench isolation (DTI) process in order to obtain an accurate AF performance. The layout and depth of DTI was optimized in order to eliminate side effects and maximize the performance even at extremely low light condition up to 1lux. In particular the AF

In order to solve the inefficient parameter update and storage issues of fine-tuning in Natural Language Generation (NLG) tasks, prompt-tuning methods have emerged as lightweight alternatives.Furthermore, efforts to reduce the gap between pre-training and fine-tuning have shown successful results in low-resource settings.As large Pre-trained Language Models (PLMs) for Program and Language Generation (PLG) tasks are constantly being developed, prompt tuning methods are necessary for the tasks.How

Code summarization is the process of automatically generating brief and informative summaries of source code to aid in software comprehension and maintenance. In this paper, we propose a novel model called READSUM, REtrieval-augmented ADaptive transformer for source code SUMmarization, that combines both abstractive and extractive approaches. Our proposed model generates code summaries in an abstractive manner, taking into account both the structural and sequential information of the input code,

Automatic processing of source code, such as code clone detection and software vulnerability detection, is very helpful to software engineers. Large pre-trained Programming Language (PL) models (such as CodeBERT, GraphCodeBERT, CodeT5, etc.), show very powerful performance on these tasks. However, these PL models are vulnerable to adversarial examples that are generated with slight perturbation. Unlike natural language, an adversarial example of code must be semantic-preserving and compilable. D

Recently, deep learning techniques have been developed for source code summarization. Most existing studies have simply adopted natural language processing techniques, because source code summarization can be considered as machine translation tasks from source code into descriptions. However, source code and its description are very different, not only in the languages of writing, but also in the purpose of writing. There is a large semantic gap between source codes in programming languages and

As pre-trained models have shown successful performance in program language processing as well as natural language processing, adversarial attacks on these models also attract attention.However, previous works on black-box adversarial attacks generated adversarial examples in a very inefficient way with simple greedy search. They also failed to find out better adversarial examples because it was hard to reduce the search space without performance loss.In this paper, we propose TABS, an efficient

In this paper, we propose a generative approach for abstractive summarization, which creates summaries based on a language model. The main goal of our paper is to generate a long sequence of words with coherent sentences by reflecting the key concepts of the original document and the characteristics of summaries. To achieve this goal, we propose an attention mechanism that uses Document Content Memory for learning the language model effectively. To evaluate its effectiveness, the proposed method

This project evaluated the corrosion damage over a five-year period of organic coated steels in automotive chassis parts during P/G and field tests using electrochemical impedance spectroscopy (EIS) in a 3.5 wt.% NaCl solution. EIS can provide both quantitative kinetic and mechanistic information about the performance of organic coating/metal systems. The difference in coating performance between the P/G and field test specimens was assessed in relation to the impedance parameters. In particular

Summary Abstractive summarization is the task of creating summaries by generating a set of novel sentences based on the information extracted from the original document, while most of summarization researches are based on extractive or compressive approaches. These approaches extract phrases from the original document and concatenate them by post‐processing and cannot truly encapsulate the contents of summaries, because they only reuse the phrases in the given document. Moreover, there are limit

Storytelling is the act of passing on what you want to tell other people as so interesting and true-to-life story. As the study in text mining progresses to express words, sentences and paragraphs as vector, it is possible to classify text and generate text using vectors. However, it has not much progressed to generate a correct flow of context and a correct grammar of context in text mining. In this paper, we propose first neural network model that learn one sentence and one vector is mapped fo