Chuan-Ming Liu | National Taipei University of Technology (original) (raw)

Papers by Chuan-Ming Liu

Agriculture

Agriculture is an important resource for the global economy, while plant disease causes devastati... more Agriculture is an important resource for the global economy, while plant disease causes devastating yield loss. To control plant disease, every country around the world spends trillions of dollars on disease management. Some of the recent solutions are based on the utilization of computer vision techniques in plant science which helps to monitor crop industries such as tomato, maize, grape, citrus, potato and cassava, and other crops. The attention-based CNN network has become effective in plant disease prediction. However, existing approaches are less precise in detecting minute-scale disease in the leaves. Our proposed Channel–Spatial segmentation network will help to determine the disease in the leaf, and it consists of two main stages: (a) channel attention discriminates diseased and healthy parts as well as channel-focused features, and (b) spatial attention consumes channel-focused features and highlights the diseased part for the final prediction process. This investigation f...

International Journal of Environmental Research and Public Health

Autistic spectrum disorder (ASD) is one of the most complex groups of neurobehavioral and develop... more Autistic spectrum disorder (ASD) is one of the most complex groups of neurobehavioral and developmental conditions. The reason is the presence of three different impaired domains, such as social interaction, communication, and restricted repetitive behaviors. Some children with ASD may not be able to communicate using language or speech. Many experts propose that continued therapy in the form of software training in this area might help to bring improvement. In this work, we propose a design of software speech therapy system for ASD. We combined different devices, technologies, and features with techniques of home rehabilitation. We used TensorFlow for Image Classification, ArKit for Text-to-Speech, Cloud Database, Binary Search, Natural Language Processing, Dataset of Sentences, and Dataset of Images with two different Operating Systems designed for Smart Mobile devices in daily life. This software is a combination of different Deep Learning Technologies and makes Human–Computer In...

Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering, 2020

The success of deep learning approaches for scene text recognition in English, Chinese and Arabic... more The success of deep learning approaches for scene text recognition in English, Chinese and Arabic language inspired us to pose a benchmark scene text recognition for Ethiopic script. To transcribe the word images to the cross bonding text, we use a segmentation free end-to-end trainable Convolutional and Recurrent Neural Network (CRNN) hybrid architecture. In the network, robust representation features from cropped word images are extracted at convolutional layer and the extracted representations features are transcribed to a sequence of labels by the recurrent layer and transcription layer. The transcription is not bounded by lexicon or word length. Due to it is effective uses to transcribe sequence-to-sequence tasks, CTC loss is applied to train the network. In order to train the proposed model, we prepare synthetic word images from Unicode fonts of Ethiopic scripts, besides the model performance is evaluated on real scene text dataset collected from different sources. The experiment result of the proposed model, shows a promising result.

EURASIA Journal of Mathematics, Science and Technology Education, 2019

The analysis of imagination has become popular in recent years because imagination is one of the ... more The analysis of imagination has become popular in recent years because imagination is one of the key components of creativity and innovation. For extracting students' implicit degrees and thought processes of imagination, we use frequent pattern mining and association rule extraction to localize the features and explain the deep meanings of imagination in the study. By our observations, these two methods may sometimes explore meaningless frequent patterns and rules on a global sparse dataset. In order to eliminate such phenomena when mining with these two methods, we use a localized feature approach called forecast with clustering and integration (FCI) to improve the drawbacks of two methods on a sparse dataset. The approach consists of two strategies. One is clustering and the other is the prediction based on integration from (1) frequent patterns, (2) association rule pruning with correlation, and (3) forecast with linear regression. The former strategy can reduce the number of samples and highlight the features of imagination and the latter strategy can prune meaningless information and predict the trend of scores from imagination input data. Experimental results show both proposed approaches can localize special features, thereby providing supervisors with meaningful information about students' degrees and thought processes of imagination.

Applied Sciences, 2020

In quantitative trading, stock prediction plays an important role in developing an effective trad... more In quantitative trading, stock prediction plays an important role in developing an effective trading strategy to achieve a substantial return. Prediction outcomes also are the prerequisites for active portfolio construction and optimization. However, the stock prediction is a challenging task because of the diversified factors involved such as uncertainty and instability. Most of the previous research focuses on analyzing financial historical data based on statistical techniques, which is known as a type of time series analysis with limited achievements. Recently, deep learning techniques, specifically recurrent neural network (RNN), has been designed to work with sequence prediction. In this paper, a long short-term memory (LSTM) network, which is a special kind of RNN, is proposed to predict stock movement based on historical data. In order to construct an efficient portfolio, multiple portfolio optimization techniques, including equal-weighted modeling (EQ), simulation modeling M...

Journal of Parallel and Distributed Computing, 2014

2009 IEEE 6th International Conference on Mobile Adhoc and Sensor Systems, 2009

ABSTRACT

Proceedings of 1994 International Conference on Parallel and Distributed Systems

The class of cographs, or complement-reducible graphs, arises naturally in many different areas o... more The class of cographs, or complement-reducible graphs, arises naturally in many different areas of applied mathematics and computer science. In this paper we present an O(n) time sequential algorithm and a parallel algorithm of O(log n) time and O(n/log n) processors on the EREW PRAM model to solve the maximum weight independent set problem on weighted cographs. Using such algorithms

2007 IEEE Wireless Communications and Networking Conference, 2007

For the cognitive radio (CR) network, a fundamental issue is how to identify the spectrum opportu... more For the cognitive radio (CR) network, a fundamental issue is how to identify the spectrum opportunities. First, each CR node determines whether there exists transmission opportunities on unlicensed bands. If not, this node will find other opportunity on licensed bands. Thus, increasing the opportunities of concurrent transmissions in unlicensed bands can reduce the overhead of wide-band sensing. In this paper, based on the carrier sensing multiple access protocol, we propose a novel concurrent transmissions MAC (CT-MAC) protocol to identify the possibility of establishing the second link in the presence of the first link in the unlicensed band. In addition to reducing the overhead of wide-band sensing, the proposed CT-MAC scheme can enhance overall throughput and is backward compatible with the IEEE 802.11 standard.

Proceedings of the 1st ACM international workshop on Performance evaluation of wireless ad hoc, sensor, and ubiquitous networks, 2004

One important issue in wireless sensor networks is how to gather sensed information in an energy ... more One important issue in wireless sensor networks is how to gather sensed information in an energy efficient way since the energy is a scarce resource in a sensor node. Many protocols have been proposed for data-gathering between wireless sensor nodes. However, most of these protocols work on static wireless sensor networks. In this paper, we provide clustering-based and time-driven protocols which minimize energy dissipation for data-gathering in wireless mobile sensor networks where the sensor nodes are capable of mobility. We first consider the node mobility when organizing clusters. Our protocols will have a sensor node select a proper clusterhead to join in order to save energy. We then consider how to elect the cluster-heads and provide two algorithms for cluster-head election. Last, we implement all the protocols and perform the experiments for evaluating the protocols.

Proceedings of the International Conference on Parallel and Distributed Systems - ICPADS, 2008

In peer-to-peer (P2P) networks, how to efficiently locate data on distributed hash tables (DHTs) ... more In peer-to-peer (P2P) networks, how to efficiently locate data on distributed hash tables (DHTs) is challenging and has attracted much research attention in recent years. Two measurements are usually considered when discussing the efficient location of data items: the path length (the hop count for resolving a lookup request on the overlay network)and the latency (the actual time period between

IEEE 60th Vehicular Technology Conference, 2004. VTC2004-Fall. 2004

Wireless sensor networks have attracted much research attention in recent years. One critical iss... more Wireless sensor networks have attracted much research attention in recent years. One critical issue in wireless sensor networks is how to gather sensed information in an energy efficient way since the energy is limited in a sensor node. Many protocols have been proposed for data-gathering or communication between wireless sensor nodes. However, most of these protocols work on static wireless sensor networks. In this paper, we provide clustering-based and time-driven protocols which minimize energy dissipation for data-gathering in wireless mobile sensor networks where the sensor nodes are capable of mobility. In many applications, the sensor nodes can move either by outside force or its mobility component. For example, the sensor nodes attached to moving objects for tracking. Our protocols consist of three major phases based on LEACH protocol: (1) cluster-head election, (2) organizing clusters, and (3) message transmission. First, we consider the node mobility when organizing clusters. Our protocols will have a sensor node select a proper clusterhead to join in order to save energy. Then, we consider how to elect the cluster-heads and provide an algorithm for cluster-head election. Last, we implement all the protocols and perform the experiments for evaluating the protocols. The simulation results show that our protocols are more energy-efficient and make the system lifetime 40-55% longer than LEACH.

The cluster-based architecture is an effective way to achieve the objective of energy efficiency ... more The cluster-based architecture is an effective way to achieve the objective of energy efficiency in wireless sensor networks. One of the critical issues in wireless sensor networks is data-gathering. In this paper, we consider the cluster-based protocol for data-gathering and explore how to elect the cluster- heads with node mobility. Two efficient distributed algorithms for cluster-head election in terms of energy consumption are provided. The proposed algorithms will make each round have the same number of cluster-heads (except the final rounds) and guarantee that each round has at least one cluster-head elected. Two mobility models, Random Walk Mobility model and Random Direction Mobility model, are considered in this paper for node mobility. Last, we implement the algorithms and perform the experiments for evaluation. The experiment results show that our cluster-head election algorithms both outperform the cluster- head election strategy used in LEACH and can make the system liv...

Sensors, 2012

This study develops and integrates an efficient knowledge-based system and a component-based fram... more This study develops and integrates an efficient knowledge-based system and a component-based framework to design an intelligent and flexible home health care system. The proposed knowledge-based system integrates an efficient rule-based reasoning model and flexible knowledge rules for determining efficiently and rapidly the necessary physiological and medication treatment procedures based on software modules, video camera sensors, communication devices, and physiological sensor information. This knowledge-based system offers high flexibility for improving and extending the system further to meet the monitoring demands of new patient and caregiver health care by updating the knowledge rules in the inference mechanism. All of the proposed functional components in this study are reusable, configurable, and extensible for system developers. Based on the experimental results, the proposed intelligent homecare system demonstrates that it can accomplish the extensible, customizable, and configurable demands of the ubiquitous healthcare systems to meet the different demands of patients and caregivers under various rehabilitation and nursing conditions.

Data broadcasting provides an effective way to disseminate information in the wireless mobile env... more Data broadcasting provides an effective way to disseminate information in the wireless mobile environment using a broadcast channel. How to provide the service of the k nearest neighbors (kNN) search using data broadcasting is studied in this paper. Given a data set D and a query point p, the kNN search finds k data points in D closest to p. By assuming that the data is indexed by an R-tree, we propose an efficient protocol for kNN search on the broadcast R-tree in terms of the tuning time which is the amount of time spent listening to the broadcast, latency which is time elapsed between issuing and termination of the query, and memory usage on the clients. We last validate the proposed protocol by extensive experiments.

The clustering architecture is essential in achieving the goal of energy efficiency for a wireles... more The clustering architecture is essential in achieving the goal of energy efficiency for a wireless sen- sor network. In general, a clustering algorithm consists of the cluster head election and the cluster member as- signment mechanism. This paper proposes an adaptive contention window (ACW)-based cluster head election mechanism. Unlike other legacy cluster head election mechanisms such as LEACH (Low Energy Adaptive Clustering Hierarchy) protocol, the proposed ACW algorithm can achieve four major goals in cluster head election for wireless sensor networks: 1) high successful probability of cluster head election, 2) appropriate number of cluster heads, 3) uniform distribution of cluster heads, and 4) equal times to be a cluster head for each sensor, simultaneously.

This paper aims to determine the optimal number of clusters in an observation area for a wireless... more This paper aims to determine the optimal number of clusters in an observation area for a wireless sensor network. We demonstrate that this goal can be achieved by a cross-layer approach from both perspectives of the power efficiency in the medium access control (MAC) layer and the coverage performance in the physical (PHY) layer.

Information Sciences, 2013

Data Broadcasting is an effective approach to provide information to a large group of clients in ... more Data Broadcasting is an effective approach to provide information to a large group of clients in ubiquitous environments. How to generate the data broadcast schedule to make the clients' average waiting time as short as possible is an important issue. In particular, when the data access pattern is dynamic and data have time constraints, such as traffic and stock information, scheduling the broadcast for such data to fulfill the requests is challenging. Since the content of the broadcast is dynamic and the request deadlines should be met, such data broadcasting is referred to as on-demand data broadcasting with time constraints. Many papers have discussed this type of data broadcasting with a single broadcast channel. In this paper, we investigate how to schedule the on-demand broadcast for the data with time constraints using multiple broadcast channels and provide two heuristics to schedule the data broadcast. The objective of the proposed heuristics is to minimize the miss rate (i.e., ratio of the number of requests missing deadlines to the number of all requests) and latency (i.e., time between issuing and termination of the request). We show that the offline version of the considered problem is NP-hard and present a competitive analysis on the proposed heuristics. More discussion about the proposed heuristics is given through extensive simulation experiments. The experimental results validate that the proposed heuristics achieve the objectives.

2021 30th Wireless and Optical Communications Conference (WOCC), 2021

Skyline is widely used in reality to solve multicriteria problems, such as environmental monitori... more Skyline is widely used in reality to solve multicriteria problems, such as environmental monitoring and business decision-making. When a data is not worse than another data on all criteria and is better than another data at least one criterion, the data is said to dominate another data. When a data item is not dominated by any other data item, this data is said to be a member of the skyline. However, as the number of criteria increases, the possibility that a data dominates another data decreases, resulting in too many members of the skyline set. To solve this kind of problem, the concept of the k-dominant skyline was proposed, which reduces the number of skyline members by relaxing the limit. The uncertainty of the data makes each data have a probability of appearing, so each data has the probability of becoming a member of the k-dominant skyline. When a new data item is added, the probability of other data becoming members of the k-dominant skyline may change. How to quickly update the k-dominant skyline for real-time applications is a serious problem. This paper proposes an effective method, Middle Indexing (MI), which filters out a large amount of irrelevant data in the uncertain data stream by sorting data specifically, so as to improve the efficiency of updating the k-dominant skyline. Experiments show that the proposed MI outperforms the existing method by approximately 13% in terms of computation time.

Life

In recent years, much research has found that dysregulation of glutarylation is associated with m... more In recent years, much research has found that dysregulation of glutarylation is associated with many human diseases, such as diabetes, cancer, and glutaric aciduria type I. Therefore, glutarylation identification and characterization are essential tasks for determining modification-specific proteomics. This study aims to propose a novel deep neural network framework based on word embedding techniques for glutarylation sites prediction. Multiple deep neural network models are implemented to evaluate the performance of glutarylation sites prediction. Furthermore, an extensive experimental comparison of word embedding techniques is conducted to utilize the most efficient method for improving protein sequence data representation. The results suggest that the proposed deep neural networks not only improve protein sequence representation but also work effectively in glutarylation sites prediction by obtaining a higher accuracy and confidence rate compared to the previous work. Moreover, e...