Takeshi Saitoh

Takeshi Saitoh
Kyushu Institute of Technology · Department of Artificial Intelligence

PhD

About

63
Publications
48,386
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
829
Citations
Additional affiliations
May 2004 - March 2010
Tottori University
Position
  • Professor (Assistant)
April 2010 - present

Publications

Publications (63)
Conference Paper
Full-text available
This paper addresses an image-based tree recognition method using shape features and color features of leaf images. Our method requires two leaf images including a front leaf image and a rear leaf image which are placed on a white background. Firstly, a leaf region is automatically extracted using a graph cuts-based method. Next, sixteen shape feat...
Article
This paper develops a complete communication support system for speech and hearing disorders using lip reading. The proposed system is a high‐level system and consists of a face detector using the Viola–Jones method, lip extraction based on the active appearance model, dynamic programming matching, automatic utterance section extraction, camera con...
Article
Full-text available
This paper develops a new functional voice activated wheelchair. Our system has two major functions. The first function is the collision avoidance function that the wheelchair avoids the wall and obstacle without voice command by using the sensor information. This function has three autonomous movements of the stop movement, avoidance movement and...
Conference Paper
Full-text available
This paper describes an automatic method for recognizing a blooming flower based on a photograph taken with a digital camera in natural scene. The problem of identifying an object against the background is known to be difficult. In this paper, we employ a photograph where the object (a blooming flower) is focused but the background is defocused. Fo...
Conference Paper
Full-text available
Aims and Objectives Collective competence of a healthcare team is key to patient safety and quality in healthcare. This presentation reports part of a UK-Japan interdisciplinary research project on emergency care team interactions, focusing on gaze behaviour. The study investigates how healthcare professionals achieve joint attention for joint acti...
Article
Grounding is a fundamental human practice for cooperation and collaboration in a joint activity, when more than two people interact. Emergency care is one such interactive situation, and whether a trauma team can efficiently establish and increment their common ground at an appropriate timing during the complex and fluid activity of emergency medic...
Chapter
This paper proposes a night time call system using a wearable camera for patients. The proposed system consists of a wearable camera, computer, relay controller, and nurse call. The user wears the wearable camera. All captured eye images are fed to the convolutional neural network to detect the pupil center. When the detected pupil center exceeds a...
Article
We are all familiar with audio speech recognition technology for interfacing with smartphones and in-car computers. However, technology that can interpret our speech signals without audio is a far greater challenge for scientists. Audio speech recognition (ASR) can only work in situations where there is little or no background noise and where speec...
Article
This paper addresses leaf images based visual tree search system called OKIRAKU Search. A user photographs an isolated leaf on a white background, and inputs the photographs to the system. The system automatically extracts the leaf region and computes shape features and color features, and searches it in the known species. The system shows the user...
Article
This paper proposes a vanishing point-based road detection method. Firstly, a vanishing point is detected using a texture-based method proposed in a recent study. After that, a histogram is generated for detecting two road borders. The road area is defined as the region between the two road borders and below the vanishing point. The experimental re...
Article
Estimating a proper location of vanishing point from a single road image without any prior known camera parameters is a challenging problem due to limited information from the input image. Most edge-based methods for vanishing point detection only work well for structured roads with clear painted lines or distinct boundaries, while they usually fai...
Article
Estimating a proper location of the road area from a single road image plays an essential role in autonomous driving systems and driver assistant systems. In this paper, we propose a new road area detection method based on texture orientations estimation and vanishing point detection. This method first estimates a vanishing point using a texture-ba...
Article
Full-text available
This paper presents a high-level real-time lip reading system that can recognize both fixed phrase and its combination. Lip reading provides an important means for realizing a communication support interface for speech handicaps. The approach is based on the Viola-Jones face detector, lip extraction based on active appearance model, automatic utter...
Article
This paper describes a real-time word lip reading system. Although our system is based on a word lip reading method already proposed, this system adapts to the user facial movements that cannot be avoided in a real-time process. With nine subjects, we obtained an average recognition rate of 89% for 14 Japanese words. Moreover, we carried out additi...
Article
Full-text available
In the field of speech recognition, the research on the auditory speech recognition which obtains high recognition accuracy tackles actively. However, in the recognition only using audio information, it is easy to be affected by the influence of surrounding noise, and there is a problem to which use environment is restricted. In recent years, the l...
Article
Full-text available
This paper presents a current sensor-based home appliance and its state recognition method for intelligent outlets. Our system has three main functions: remote control, monitoring, and power supply schedule management. This research focuses particular on the monitoring function. To recognize the appliance and the state of the appliance, we extract...
Conference Paper
Full-text available
This paper proposes the mobile robot embedded two functions, the following function and returning function in the indoor environment with monocular camera. In the following mode, the robot follows the target object such as the person who walks in front of robot, and runs until reaching his destination. To follow him, the region extraction method is...
Conference Paper
Full-text available
This paper develops a functional voice activated wheelchair. Various interfaces to control powered wheelchair are proposed. Since the voice is the most natural communication ways for person, our study pays attention to speech recognition. The user controls the wheelchair by the interactive operation. The wheelchair does not act based on false speec...
Conference Paper
Full-text available
This paper proposes the mobile robot consisted of following function and returning function with monocular camera. In the following function, the robot follows the target object such as the person who walks in front of robot, and moves until reaching his destination. To follow him, the region extraction method is applied. Furthermore, the robot rec...
Conference Paper
Full-text available
The traditional researches targeted at only one language, and there is no research to refer the language and recognition method. Moreover, a lot of model-based methods use only an external lip or intraoral region, and tooth or tongue region is not reflected to the feature. This paper describes analysis of efficient lip reading method for various la...
Article
Full-text available
This paper describes the novel transducer which rec-ognizes lip motion of the word utterance and output the recognized word as the voice message. Though, the base method of our transducer is speaker dependent word lip reading, our system is adopted the facial move-ment of the user. For the developed prototype system, the robustness was evaluated to...
Article
Full-text available
Recently, the increasing bird injury is one of the social problems. To solve this problem, we set our goal to develop the monitoring system. As the first step, we focus the bird tracking and flapping motion recognition both techniques are core technologies of our goal. It is found that Mean Shift tracking based on both the color model and flow mode...
Article
Full-text available
SUMMARY This paper describes a recognition method of Japanese single sounds for application to lip reading. Related researches investi- gated only five or ten sounds. In this paper, experiments were conducted for 45 Japanese single sounds by classifying them into five vowels category, ten consonants category, and 45 sounds category. We obtained rec...
Chapter
Full-text available
This paper proposed the appearance based method for detecting two boundary lines between the wall and corridor and the obstacle region through the image processing based on monocular vision. Moreover, the proposed method was implemented in the wheelchair based indoor mobile robot. The developed robot moved at the center of the corridor, and it work...
Article
Full-text available
This paper presents the indoor mobile robot that moves automatically without needing environment informa-tion beforehand while recognizing the frontal surrounding en-vironment with only one general camera. Based on the frontal image, the robot detects two boundary lines, some obstacle regions, and a moving direction. When the obstacle is de-tected,...
Article
Full-text available
SUMMARY This paper analyses the features required to efficiently rec- ognize five Japanese vowels for lip-reading. Various features, such as shape and radius, are calculated from the lip region and fed to the k Nearest Neighbor method. We calculated 15 feature sets and found that the fea- ture set including the area and aspect ratio of the mouth ca...
Conference Paper
Full-text available
A system for center following in the corridor using monocular vision is presented. The robot detects a vanishing point and obstacle region from frontal view image. The vanishing point is found by two boundary lines of the wall and the corridor. The obstacle region is detected from corridor region using color histogram. The experiments are performed...
Conference Paper
Full-text available
In order to assist physically handicapped persons, we developed a voice controlled wheelchair. The user can control the wheelchair by voice commands, such as "susume (run forward)" in Japanese. A grammar-based recognition parser named "Julian" is used in our system. Three type commands, the basic reaction command, the short moving reaction command,...
Conference Paper
Full-text available
This paper describes the novel interface equipped intelligent wheelchair. The proposed interface is using visual oral motion. In our system, the user moves his mouth open and close to run the wheelchair. To achieve detection of mouth region and analysis mouth motion in real-time, we proposed mouth cavity region detection method. We carried out the...
Conference Paper
Full-text available
This paper describes an intelligent wheelchair with a novel interface which uses visual oral motion. The wheelchair runs in various lighting environments, traditional mouth region detection method did not consider these conditions. To achieve detection of mouth region and analysis mouth motion in real-time and in various lighting environments, we p...
Conference Paper
Full-text available
Recently, speech, especially word recognition using visual information, has attracted significantly interest. In word recognition, the target is not just a word, but also a vowel. However, since vowel frames do not contain many phonemes, vowel recognition rate is less than that of the word. Some research has been done on vowel recognition. This pap...
Conference Paper
Full-text available
This paper describes Japanese phone recognition for lip reading based on a novel feature called trajectory feature to obtain high recognition rate. Trajectory fea- ture is a time change of two mouth shape features ex- pressed as a two-dimensional trajectory of the lip mo- tion. The most similar trajectory in a database which is compared with the ta...
Article
This paper describes an automatic detection and tracking method for unknown object from unclear background. Our method consists of two processes, the detection process and the tracking process. The former process is the background subtraction method and the block-based clustering. The latter process is with mean shift method. To detect an accurate...
Article
Full-text available
The visual information doesn't have an influence with the noise in comparison with the auditory information. However, lip-reading is earlier technology than speech recognition, there is a problem that recognition rate is low. This paper proposes a novel feature to obtain high recognition rate for lip-reading. The proposed lip-reading method is as f...
Conference Paper
Full-text available
This paper describes about real-time object tracking method based on mean shift tracking algorithm. The mean shift tracking algorithm is an efficient technique for tracking object through an image. The original mean shift tracking is proposed to apply the color image based on the color distribution. A near-infrared camera is used with surveillance...
Conference Paper
Full-text available
This paper describes the lip reading method using video and thermal images. We pay attention to the lip movement. The density of the thermal image expresses the surface temperature, and it is difficult to detect the lip position automatically. Though the video and the thermal images take in the same time, these don't synchronized in time and spatia...
Conference Paper
Full-text available
This paper describes the development of the monocular autonomy following vehicle which for reduce a helper burden. The system is composed by a CCD camera, a FPGA board and some control circuits. Our idea is that the distance between the system and the target human is related to the size of the human region in the camera image. Based on this idea, w...
Article
Full-text available
This paper proposed a method in which location of a person is estimation with one active camera in real-time. The approach taken in this paper is as follows. First, we detect the person region with proposed method which is the region detection method based on short time distribution of the density to avoid the influence of the camera movement. Then...
Article
SUMMARY In this paper, we propose a method for extracting an object region from an image. First, we investigate the properties of the Intelligent Scissors (IS) method, which has been proposed as a method for extracting contours. As a result, we verify that the IS method's route search according to total cost requires multiple seed points to extract...
Conference Paper
Full-text available
This paper describes a model-based method for detecting lip region from image sequences. Our approach is by Sampled Active Contour Model (S-ACM). The original S-ACM has the problem which can’t expand. To overcome this problem, we propose the elastic S-ACM. Moreover, based on the extracted lip contour, the effective delta radius features are fed to...
Article
SUMMARY Recently, high-resolution multislice CT images have been used in the diagnosis of hepatic cancer. In the conven- tional method of detecting hepatic cancer on the basis of three-dimensional images, it is difficult to detect cancers with a three-dimensional structure. This paper proposes a method of detection of liver cancer on the basis of t...
Article
Method of quartz crystal oscillator of odor sensors is drawing public attention recently. In this paper, we modulated gas flow used a magnetic valve in order to improve the rate of discrimination. Furthermore, we change into urethane resin the Film applied to a quartz crystal oscillator, and also report the result which used and discriminated the t...
Article
In this paper, we propose a small-scale remote video conference system, and this paper describes a method for detecting a speaker among the participants based on the voice and the image processing of lip. At first, its method detects the direction of sound source and turns to camera to a speaker's direction. Then, the lip shape which has the most a...
Article
SUMMARY With the rapid progress of CT (Computed Tomogra- phy) scanners, four-phase CT images with resolutions as high as 1 mm have started to be used for diagnosing liver diseases. The first-, second-, third-, and fourth-phase CT images correspond to before dye injection, the early stage, the full stage, and the wash-out stage of the injected dye....
Article
In this paper we propose an automatic recognition system for wild flowers. Two photos, one of the flower and one of the leaves taken from directly above or at a close angle of a single wild flower, were used as a single set. The objective (flower, leaf) is extracted from each image using a clustering method, then recognition is performed using a pi...
Conference Paper
Full-text available
This paper presents a new method for automatically extracting an object region from a photograph based upon a well-known method “Intelligent Scissors” (IS). For our application, it will be shown that (1) the cost should not be based on accumulated cost adopted by IS but rather on average cost and (2) only a few past pixels are needed for deciding t...
Article
Texture synthesis based upon a sample image or template has attracted much attention recently. This paper describes a new blending algorithm for synthesizing a transient texture based upon two or more templates that gives an impression of a gradual transition from one template to another. Based upon a pair combination out of four templates, we synt...
Conference Paper
Full-text available
Recently four phases CT has been widely used for diagnosing liver diseases, especially cancer. This paper describes an automatic method for segmenting the liver region from third phase abdominal CT. First, blood vessels in the liver are extracted with a threshold. Then morphological dilation enables us to define an approximate liver region, which i...
Conference Paper
Full-text available
This paper describes an automatic method for recognizing wild flowers. Recognition requires two pictures; a frontal flower image and a leaf image taken by a digital camera. Seventeen features, eight from the flower and also nine from the leaf are fed to a neural network. We collected 20 pairs of pictures from 16 wild flowers in the fields around ou...
Article
Full-text available
This paper presents the current sensor based non-intrusive appliance recognition method for intelligent outlet. Our system has two main functions; one is the remote control function of power supply through the Internet. The other is monitoring function observe the state of appliance. In this pa-per, the monitor function is especially focused. To re...

Network

Cited By