• Title/Summary/Keyword: Multi-Stream

Search Result 411, Processing Time 0.029 seconds

Two-Stream Convolutional Neural Network for Video Action Recognition

  • Qiao, Han;Liu, Shuang;Xu, Qingzhen;Liu, Shouqiang;Yang, Wanggan
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • v.15 no.10
    • /
    • pp.3668-3684
    • /
    • 2021
  • Video action recognition is widely used in video surveillance, behavior detection, human-computer interaction, medically assisted diagnosis and motion analysis. However, video action recognition can be disturbed by many factors, such as background, illumination and so on. Two-stream convolutional neural network uses the video spatial and temporal models to train separately, and performs fusion at the output end. The multi segment Two-Stream convolutional neural network model trains temporal and spatial information from the video to extract their feature and fuse them, then determine the category of video action. Google Xception model and the transfer learning is adopted in this paper, and the Xception model which trained on ImageNet is used as the initial weight. It greatly overcomes the problem of model underfitting caused by insufficient video behavior dataset, and it can effectively reduce the influence of various factors in the video. This way also greatly improves the accuracy and reduces the training time. What's more, to make up for the shortage of dataset, the kinetics400 dataset was used for pre-training, which greatly improved the accuracy of the model. In this applied research, through continuous efforts, the expected goal is basically achieved, and according to the study and research, the design of the original dual-flow model is improved.

1H*-tree: An Improved Data Cube Structure for Multi-dimensional Analysis of Data Streams (1H*-tree: 데이터 스트림의 다차원 분석을 위한 개선된 데이터 큐브 구조)

  • XiangRui Chen;YuXiang Cheng;Yan Li;Song-Sun Shin;Dong-Wook Lee;Hae-Young Bae
    • Proceedings of the Korea Information Processing Society Conference
    • /
    • 2008.11a
    • /
    • pp.332-335
    • /
    • 2008
  • In this paper, based on H-tree, which is proposed as the basic data cube structure for multi-dimensional data stream analysis, we have done some analysis. We find there are a lot of redundant nodes in H-tree, and the tree-build method can be improved for saving not only memory, but also time used for inserting tuples. Also, to facilitate more fast and large amount of data stream analysis, which is very important for stream research, H*-tree is designed and developed. Our performance study compare the proposed H*-tree and H-tree, identify that H*-tree can save more memory and time during inserting data stream tuples.

Design of Sensor Middleware Architecture on Multi Level Spatial DBMS with Snapshot (스냅샷을 가지는 다중 레벨 공간 DBMS를 기반으로 하는 센서 미들웨어 구조 설계)

  • Oh, Eun-Seog;Kim, Ho-Seok;Kim, Jae-Hong;Bae, Hae-Young
    • Journal of Korea Spatial Information System Society
    • /
    • v.8 no.1 s.16
    • /
    • pp.1-16
    • /
    • 2006
  • Recently, human based computing environment for supporting users to concentrate only user task without sensing other changes from users is being progressively researched and developed. But middleware deletes steream data processed for reducing process load of massive information from RFID sensor in this computing. So, this kind of middleware have problems when user demands probability or statistics needed for data warehousing or data mining and when user demands very important stream data repeatedly but already discarded in the middleware every former time. In this paper, we designs Sensor Middleware Architecture on Multi Level Spatial DBMS with Snapshot and manage repeatedly required stream datas to solve reusing problems of historical stream data in current middleware. This system uses disk databse that manages historical stream datas filtered in middleware for requiring services using historical stream information as data mining or data warehousing from user, and uses memory database that mamages highly reuseable data as a snapshot when stream data storaged in disk database has high reuse frequency from user. For the more, this system processes memory database management policy in a cycle to maintain high reusement and rapid service for users. Our paper system solves problems of repeated requirement of stream datas, or a policy decision service using historical stream data of current middleware. Also offers variant and rapid data services maintaining high data reusement of main memory snapshot datas.

  • PDF

Assessment of stream water quality and pollutant discharge loads affected by recycled irrigation in an agricultural watershed using HSPF and a multi-reservoir model (HSPF와 다중 저류지 모형을 이용한 농업지역 순환관개에 의한 하천 수질 및 배출부하 영향 분석)

  • Kyoung-Seok Lee;Dong Hoon Lee;Youngmi Ahn;Joo-Hyon Kang
    • Journal of Wetlands Research
    • /
    • v.25 no.4
    • /
    • pp.297-305
    • /
    • 2023
  • The recycled irrigation is a type of irrigation that uses downstream water to fulfill irrigation demand in the upstream agricultural areas; the used irrigation water returns back to the downstream. The recycled irrigation is advantageous for securing irrigation water for plant growth, but the returned water typically contains high levels of nutrients due to excess nutrients inputs during the agricultural activities, potentially deteriorating stream water quality. Therefore, quantitative assessment on the effect of the recycled irrigation on the stream water quality is required to establish strategies for effective irrigation water supply and water quality management. For this purpose, a watershed model is generally used; however no functions to simulate the effects of the recycled irrigation are provided in the existing watershed models. In this study, we used multi-reservoir model coupled with the Hydrological Simulation Program-Fortran (HSPF) to estimate the effect of the recycled irrigation on the stream water quality. The study area was the Gwangok stream watershed, a subwatershed of Gyeseong stream watershed in Changnyeong county, Gyeongsangnam-do. The HSPF model was built, calibrated, and used to produce time series data of flow and water quality, which were used as hypothetical observation data to calibrate the multi-reservoir model. The calibrated multi-reservoir model was used for simulating the recycled irrigation. In the multi-reservoir model, the Gwangok watershed consisted of two subsystems, irrigation and the Gwangok stream, and the reactions (plant uptake, adsorption, desorption, and decay) within each subsystem, and fluxes of water and materials between the subsystems, were modeled. Using the developed model, three scenarios with different combinations of the operating conditions of the recycled irrigation were evaluated for their effects on the stream water quality.

Optimizing Multi-way Join Query Over Data Streams (데이타 스트림에서의 다중 조인 질의 최적화 방법)

  • Park, Hong-Kyu;Lee, Won-Suk
    • Journal of KIISE:Databases
    • /
    • v.35 no.6
    • /
    • pp.459-468
    • /
    • 2008
  • A data stream which is a massive unbounded sequence of data elements continuously generated at a rapid rate. Many recent research activities for emerging applications often need to deal with the data stream. Such applications can be web click monitoring, sensor data processing, network traffic analysis. telephone records and multi-media data. For this. data processing over a data stream are not performed on the stored data but performed the newly updated data with pre-registered queries, and then return a result immediately or periodically. Recently, many studies are focused on dealing with a data stream more than a stored data set. Especially. there are many researches to optimize continuous queries in order to perform them efficiently. This paper proposes a query optimization algorithm to manage continuous query which has multiple join operators(Multi-way join) over data streams. It is called by an Extended Greedy query optimization based on a greedy algorithm. It defines a join cost by a required operation to compute a join and an operation to process a result and then stores all information for computing join cost and join cost in the statistics catalog. To overcome a weak point of greedy algorithm which has poor performance, the algorithm selects the set of operators with a small lay, instead of operator with the smallest cost. The set is influenced the accuracy and execution time of the algorithm and can be controlled adaptively by two user-defined values. Experiment results illustrate the performance of the EGA algorithm in various stream environments.

Multi-focus Image Fusion using Fully Convolutional Two-stream Network for Visual Sensors

  • Xu, Kaiping;Qin, Zheng;Wang, Guolong;Zhang, Huidi;Huang, Kai;Ye, Shuxiong
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • v.12 no.5
    • /
    • pp.2253-2272
    • /
    • 2018
  • We propose a deep learning method for multi-focus image fusion. Unlike most existing pixel-level fusion methods, either in spatial domain or in transform domain, our method directly learns an end-to-end fully convolutional two-stream network. The framework maps a pair of different focus images to a clean version, with a chain of convolutional layers, fusion layer and deconvolutional layers. Our deep fusion model has advantages of efficiency and robustness, yet demonstrates state-of-art fusion quality. We explore different parameter settings to achieve trade-offs between performance and speed. Moreover, the experiment results on our training dataset show that our network can achieve good performance with subjective visual perception and objective assessment metrics.

A development of rating-curve using Bayesian Multi-Segmented model (Bayesian 기반 Multi-Segmented 곡선식을 활용한 수위-유량 곡선의 불확실성 분석)

  • Kim, Jin-Young;Kim, Jin-Guk;Lee, Jae Chul;Kwon, Hyun-Han
    • Journal of Korea Water Resources Association
    • /
    • v.49 no.3
    • /
    • pp.253-262
    • /
    • 2016
  • A Rating curve is a regression equation of discharge versus stage for a given point on a stream where the stream discharge is measured across the stream channel with a stage and discharge measurement. The curve is generally used to calculate discharge based on the stage. However, the existing approach showed problems in terms of estimating uncertainty associated with regression parameters including the separation parameter for low and high flow. In this regard, this study aimed to develop a new method for the aforementioned problems based on Bayesian approach, which can better estimate the parameter and its uncertainty. In addition, this study used a Bayesian Multi-Segmented (Bayesian M-S) model which is provided a comparison between the existing and proposed scheme. The proposed model showed better results for the parameter estimation than the existing approach, and provided better performance in terms of estimating uncertainty range.

The Method for Extracting Meaningful Patterns Over the Time of Multi Blocks Stream Data (시간의 흐름과 위치 변화에 따른 멀티 블록 스트림 데이터의 의미 있는 패턴 추출 방법)

  • Cho, Kyeong-Rae;Kim, Ki-Young
    • KIPS Transactions on Computer and Communication Systems
    • /
    • v.3 no.10
    • /
    • pp.377-382
    • /
    • 2014
  • Analysis techniques of the data over time from the mobile environment and IoT, is mainly used for extracting patterns from the collected data, to find meaningful information. However, analytical methods existing, is based to be analyzed in a state where the data collection is complete, to reflect changes in time series data associated with the passage of time is difficult. In this paper, we introduce a method for analyzing multi-block streaming data(AM-MBSD: Analysis Method for Multi-Block Stream Data) for the analysis of the data stream with multiple properties, such as variability of pattern and large capacitive and continuity of data. The multi-block streaming data, define a plurality of blocks of data to be continuously generated, each block, by using the analysis method of the proposed method of analysis to extract meaningful patterns. The patterns that are extracted, generation time, frequency, were collected and consideration of such errors. Through analysis experiments using time series data.

열추적자를 이용한 지하수-하천수 혼합대 연구

  • Kim Gu-Yeong;Jeon Cheol-Min;Kim Tae-Hui;Seong Hyeon-Jeong;O Jun-Ho;Kim Yong-Je;Jeong Jae-Hun;Park Seung-Gi
    • Proceedings of the Korean Society of Soil and Groundwater Environment Conference
    • /
    • 2006.04a
    • /
    • pp.277-281
    • /
    • 2006
  • A study on stream-groundwater exchange was performed using head and temperature data of stream water, streambed, and groundwater. Groundwater level and temperature were obtained from multi-depth monitoring wells in small-scale watershed. In the summer time, time series of temperatrue data at streambed and groundwater were monitored for three months. In the winter time, we measured the temperature gradient between stream water and streambed. The observed data showed three typical types of temperature characteristics. First, the temperature of streambed was lower than that of stream water; second, the temperature of streambed and stream water was similar; and last, the temperature of streambed was higher than that of stream water. The interconnections between the stream and the streambed were not homogeneously distributed due to weakly developed sediments and heterogeneous bedrock exposed as bed of the stream. The temperature data may be used in formal solutions of the inverse problems to estimate groundwater flow and hydraulic conductivity.

  • PDF

Channel Transition Analysis of Smart HLS with Dynamic Single Buffering Scheme (동적 단일 버퍼링 기법을 적용한 스마트 HLS의 채널변경 분석)

  • Kim, Chong-il;Kang, Min-goo;Kim, Dong-hyun;Kim, In-ki;Han, Kyung-sik
    • Journal of Internet Computing and Services
    • /
    • v.17 no.6
    • /
    • pp.9-15
    • /
    • 2016
  • In this paper, we propose a smart HLS(HTTP Live Stream) platform with dynamic single buffering for the best transmission of adaptive video bit-rates. This smart HLS can optimizes the channel transition zapping-time with the monitoring of bandwidth between HLS server and OTT(Over The Top) client. This platform is designed through the control of video stream due to proper multi-bitrates and bandwidths. This proposed OTT can decode the live and VOD(Video On Demand) videos with the buffering of optimumal bitrate. And, the HLS can be cooperated with a smart OTT, and segmented for the m3u8 files of H.265 MPEG-2 TS(Transport Stream) videos. As a resullt, this single buffer based smart OTT can transmit optimal videos with the maximum data buffering according to the adaptive bit-rate depending on the network bandwidth efficiency and the decoded VOD video, too.