• Title/Summary/Keyword: DNN 모델

Search Result 192, Processing Time 0.026 seconds

Predicting Program Code Changes Using a CNN Model (CNN 모델을 이용한 프로그램 코드 변경 예측)

  • Kim, Dong Kwan
    • Journal of the Korea Convergence Society
    • /
    • v.12 no.9
    • /
    • pp.11-19
    • /
    • 2021
  • A software system is required to change during its life cycle due to various requirements such as adding functionalities, fixing bugs, and adjusting to new computing environments. Such program code modification should be considered as carefully as a new system development becase unexpected software errors could be introduced. In addition, when reusing open source programs, we can expect higher quality software if code changes of the open source program are predicted in advance. This paper proposes a Convolutional Neural Network (CNN)-based deep learning model to predict source code changes. In this paper, the prediction of code changes is considered as a kind of a binary classification problem in deep learning and labeled datasets are used for supervised learning. Java projects and code change logs are collected from GitHub for training and testing datasets. Software metrics are computed from the collected Java source code and they are used as input data for the proposed model to detect code changes. The performance of the proposed model has been measured by using evaluation metrics such as precision, recall, F1-score, and accuracy. The experimental results show the proposed CNN model has achieved 95% in terms of F1-Score and outperformed the multilayer percept-based DNN model whose F1-Score is 92%.

Compromised feature normalization method for deep neural network based speech recognition (심층신경망 기반의 음성인식을 위한 절충된 특징 정규화 방식)

  • Kim, Min Sik;Kim, Hyung Soon
    • Phonetics and Speech Sciences
    • /
    • v.12 no.3
    • /
    • pp.65-71
    • /
    • 2020
  • Feature normalization is a method to reduce the effect of environmental mismatch between the training and test conditions through the normalization of statistical characteristics of acoustic feature parameters. It demonstrates excellent performance improvement in the traditional Gaussian mixture model-hidden Markov model (GMM-HMM)-based speech recognition system. However, in a deep neural network (DNN)-based speech recognition system, minimizing the effects of environmental mismatch does not necessarily lead to the best performance improvement. In this paper, we attribute the cause of this phenomenon to information loss due to excessive feature normalization. We investigate whether there is a feature normalization method that maximizes the speech recognition performance by properly reducing the impact of environmental mismatch, while preserving useful information for training acoustic models. To this end, we introduce the mean and exponentiated variance normalization (MEVN), which is a compromise between the mean normalization (MN) and the mean and variance normalization (MVN), and compare the performance of DNN-based speech recognition system in noisy and reverberant environments according to the degree of variance normalization. Experimental results reveal that a slight performance improvement is obtained with the MEVN over the MN and the MVN, depending on the degree of variance normalization.

A Comparative Study of Machine Learning Algorithms Based on Tensorflow for Data Prediction (데이터 예측을 위한 텐서플로우 기반 기계학습 알고리즘 비교 연구)

  • Abbas, Qalab E.;Jang, Sung-Bong
    • KIPS Transactions on Computer and Communication Systems
    • /
    • v.10 no.3
    • /
    • pp.71-80
    • /
    • 2021
  • The selection of an appropriate neural network algorithm is an important step for accurate data prediction in machine learning. Many algorithms based on basic artificial neural networks have been devised to efficiently predict future data. These networks include deep neural networks (DNNs), recurrent neural networks (RNNs), long short-term memory (LSTM) networks, and gated recurrent unit (GRU) neural networks. Developers face difficulties when choosing among these networks because sufficient information on their performance is unavailable. To alleviate this difficulty, we evaluated the performance of each algorithm by comparing their errors and processing times. Each neural network model was trained using a tax dataset, and the trained model was used for data prediction to compare accuracies among the various algorithms. Furthermore, the effects of activation functions and various optimizers on the performance of the models were analyzed The experimental results show that the GRU and LSTM algorithms yields the lowest prediction error with an average RMSE of 0.12 and an average R2 score of 0.78 and 0.75 respectively, and the basic DNN model achieves the lowest processing time but highest average RMSE of 0.163. Furthermore, the Adam optimizer yields the best performance (with DNN, GRU, and LSTM) in terms of error and the worst performance in terms of processing time. The findings of this study are thus expected to be useful for scientists and developers.

Multiaspect-based Active Sonar Target Classification Using Deep Belief Network (DBN을 이용한 다중 방위 데이터 기반 능동소나 표적 식별)

  • Kim, Dong-wook;Bae, Keun-sung;Seok, Jong-won
    • Journal of the Korea Institute of Information and Communication Engineering
    • /
    • v.22 no.3
    • /
    • pp.418-424
    • /
    • 2018
  • Detection and classification of underwater targets is an important issue for both military and non-military purposes. Recently, many performance improvements are being reported in the field of pattern recognition with the development of deep learning technology. Among the results, DBN showed good performance when used for pre-training of DNN. In this paper, DBN was used for the classification of underwater targets using active sonar, and the results are compared with that of the conventional BPNN. We synthesized active sonar target signals using 3-dimensional highlight model. Then, features were extracted based on FrFT. In the single aspect based experiment, the classification result using DBN was improved about 3.83% compared with the BPNN. In the case of multi-aspect based experiment, a performance of 95% or more is obtained when the number of observation sequence exceeds three.

Performance comparison evaluation of speech enhancement using various loss functions (다양한 손실 함수를 이용한 음성 향상 성능 비교 평가)

  • Hwang, Seo-Rim;Byun, Joon;Park, Young-Cheol
    • The Journal of the Acoustical Society of Korea
    • /
    • v.40 no.2
    • /
    • pp.176-182
    • /
    • 2021
  • This paper evaluates and compares the performance of the Deep Nerual Network (DNN)-based speech enhancement models according to various loss functions. We used a complex network that can consider the phase information of speech as a baseline model. As the loss function, we consider two types of basic loss functions; the Mean Squared Error (MSE) and the Scale-Invariant Source-to-Noise Ratio (SI-SNR), and two types of perceptual-based loss functions, including the Perceptual Metric for Speech Quality Evaluation (PMSQE) and the Log Mel Spectra (LMS). The performance comparison was performed through objective evaluation and listening tests with outputs obtained using various combinations of the loss functions. Test results show that when a perceptual-based loss function was combined with MSE or SI-SNR, the overall performance is improved, and the perceptual-based loss functions, even exhibiting lower objective scores showed better performance in the listening test.

Comparison of Performance of LSTM and EEMD based PM10 Prediction Model (LSTM과 EEMD 기반의 미세먼지 농도 예측 모델 성능 비교)

  • Jung, Yong-jin;Oh, Chang-heon
    • Proceedings of the Korean Institute of Information and Commucation Sciences Conference
    • /
    • 2022.05a
    • /
    • pp.510-512
    • /
    • 2022
  • Various studies are being conducted to improve the accuracy of fine dust, but there is a problem that deep learning models are not well learned due to various characteristics according to the concentration of fine dust. This paper proposes an EEMD-based fine dust concentration prediction model to decompose the characteristics of fine dust concentration and reflect the characteristics. After decomposing the fine dust concentration through EEMD, the final fine dust concentration value is derived by ensemble of the prediction results according to the characteristics derived from each. As a result of the model's performance evaluation, 91.7% of the fine dust concentration prediction accuracy was confirmed.

  • PDF

Study on the Selection of Optimal Operation Position Using AI Techniques (인공지능 기법에 의한 최적 운항자세 선정에 관한 연구)

  • Dong-Woo Park
    • Journal of the Korean Society of Marine Environment & Safety
    • /
    • v.29 no.6
    • /
    • pp.681-687
    • /
    • 2023
  • The selection technique for optimal operation position selection technique is used to present the initial bow and stern draft with minimum resistance, for achievingthat is, the optimal fuel consumption efficiency at a given operating displacement and speed. The main purpose of this studypaper is to develop a program to select the optimal operating position with maximum energy efficiency under given operating conditions based on the effective power data of the target ship. This program was written as a Python-based GUI (Graphic User Interface) usingbased on artificial intelligence techniques sucho that ship owners could easily use the GUIit. In the process, tThe introduction of the target ship, the collection of effective power data through computational fluid dynamics (CFD), the learning method of the effective power model using deep learning, and the program for presenting the optimal operation position using the deep neural network (DNN) model were specifically explained. Ships are loaded and unloaded for each operation, which changes the cargo load and changes the displacement. The shipowners wants to know the optimal operating position with minimum resistance, that is, maximum energy efficiency, according to the given speed of each displacement. The developed GUI can be installed on the ship's tablet PC and application and used to determineselect the optimal operating position.

Lightweight Network for Multi-exposure High Dynamic Range Imaging (다중 노출 High Dynamic Range 이미징을 위한 경량화 네트워크)

  • Lee, Keuntek;Cho, Nam Ik
    • Proceedings of the Korean Society of Broadcast Engineers Conference
    • /
    • fall
    • /
    • pp.70-73
    • /
    • 2021
  • 최근 영상 및 비디오 분야에 심층 신경망(DNN, Deep Neural Network)을 사용한 연구가 다양하게 진행됨에 따라 High Dynamic Range (HDR) 이미징 기술에서도 기존의 방법들 보다 우수한 성능을 보이는 심층 신경망 모델들이 등장하였다. 하지만, 심층 신경망을 사용한 방법은 큰 연산량과 많은 GPU 메모리를 사용한다는 문제점이 존재하며, 이는 심층 신경망 기반 기술들의 현실 적용 가능성에 제한이 되고 있다. 이에 본 논문에서는 제한된 연산량과 GPU 메모리 조건에서도 사용 가능한 다중 노출 HDR 경량화 심층 신경망을 제안한다. Kalantari Dataset에 대해 기존 HDR 모델들과의 성능 평가를 진행해 본 결과, PSNR-µ와 PSNR-l 수치에서 GPU 메모리 사용량 대비 우수한 성능을 보임을 확인하였다.

  • PDF

Performance Evaluation of FPN-Attention Layered Model for Improving Visual Explainability of Object Recognition (객체 인식 설명성 향상을 위한 FPN-Attention Layered 모델의 성능 평가)

  • Youn, Seok Jun;Cho, Nam Ik
    • Proceedings of the Korean Society of Broadcast Engineers Conference
    • /
    • 2022.06a
    • /
    • pp.1311-1314
    • /
    • 2022
  • DNN을 사용하여 객체 인식 과정에서 객체를 잘 분류하기 위해서는 시각적 설명성이 요구된다. 시각적 설명성은 object class에 대한 예측을 pixel-wise attribution으로 표현해 예측 근거를 해석하기 위해 제안되었다, Scale-invariant한 특징을 제공하도록 설계된 pyramidal features 기반 backbone 구조는 object detection 및 classification 등에서 널리 쓰이고 있으며, 이러한 특징을 갖는 feature pyramid를 trainable attention mechanism에 적용하고자 할 때 계산량 및 메모리의 복잡도가 증가하는 문제가 있다. 본 논문에서는 일반적인 FPN에서 객체 인식 성능과 설명성을 높이기 위한 피라미드-주의집중 계층네트워크 (FPN-Attention Layered Network) 방식을 제안하고, 실험적으로 그 특성을 평가하고자 한다. 기존의 FPN만을 사용하였을 때 객체 인식 과정에서 설명성을 향상시키는 방식이 객체 인식에 미치는 정도를 정량적으로 평가하였다. 제안된 모델의 적용을 통해 낮은 computing 오버헤드 수준에서 multi-level feature를 고려한 시각적 설명성을 개선시켜, 결괴적으로 객체 인식 성능을 향상 시킬 수 있음을 실험적으로 확인할 수 있었다.

  • PDF

Accelerating Neural Network Inference using SIMD in Resource-Constrained Environments (자원 제약 환경에서 SIMD 를 활용한 신경망 연산 가속)

  • Se-Hyeon Jeong;Gi-won Kang;Yun-Seo Lee;Bon-Wook Gu;Jeong-Min Hwang;Hyunyoung Oh
    • Annual Conference of KIPS
    • /
    • 2024.10a
    • /
    • pp.50-51
    • /
    • 2024
  • 본 연구는 자원 제약적 임베디드 시스템에서 신경망 연산의 효율성을 극대화하기 위해 SIMD(Single Instruction Multiple Data) 기술을 활용한 최적화 기법을 제안한다. 기존 연구들이 주로 합성곱 연산에 집중된 것과 달리, 본 연구는 신경망의 전체 연산 구간에 SIMD 최적화를 적용하고, 범용 DNN 프레임워크인 Darknet 을 기반으로 다양한 모델에 적용 가능한 방법론을 적용하였다. Raspberry Pi 3B+를 테스트베드로 활용하여 다양한 CNN 모델에 대한 성능 평가를 수행하였으며, 최대 55.2%의 성능 향상을 달성하였다. 또한, SIMD 레지스터 활용도와 연산 속도 간의 상관관계를 분석하여 최적의 구현 전략을 도출하였다.