JongSuk Choi

Near-time-optimal trajectory planning for wheeled mobile robots with translational and rotational sections

We derive a near-time-optimal trajectory for wheeled mobile robots (WMRs) satisfying the following: 1) initial and final postures/velocities; and 2) battery voltage and armature current constraints, under assumptions of simplified dynamics and constant translational/rotational velocity sections. We use a simplified dynamic model for WMRs, neglecting inductances of motor armatures and divide our trajectory generation algorithm for cornering motion into three sections. We specify a path-deviation requirement for obstacle avoidance. Transforming dynamics into uncorrelated form with regard to translational and rotational velocities, we make extreme control possible. By separating the rotational section with translational sections and determining the velocity scale factor a near-time-optimal trajectory can be obtained. Simulation results along with inverse control of path-following are given to validate the generated trajectory.

Advanced indoor localization using ultrasonic sensor and digital compass

Several kinds of indoor localization systems have been developed for estimating objects position lately. They are classified according to the environment, the purpose of application, the media type and so on. In this paper, we propose an improved beacon system, having the ultrasonic media type, which can be used for indoor localization; Moreover it can estimate the direction of a mobile object by using a digital compass. However it is very hard for the indoor localization system using ultrasonic sensors to achieve much highly precise localization for mobile objects since the ultrasonic sensor is highly sensitive to noises and external shocks. In order to overcome this problem, we adopt a bandpass-filter to protect ultrasonic sensor from another frequency generated by noise and external shocks. In addition, the multi-modulation of ultrasonic sensors - using two ultrasonic sensors with different frequencies - improves the sampling rate of this system. And we present the application of Unscented Kalman Filter to improve the localization performance of our indoor mobile localization system.

Reliable detection of sound's direction for human robot interaction

In this paper, we propose reliable detection of sounds direction for human robot interaction. Compared with previous researches, this system comprises simpler algorithm and a proposed nonlinear amplifier which has advantages to increase a detectible distance of sound signal in spite of simple circuits. Moreover, we propose the new performance index using cross-correlation to make reliable detection of sounds direction. Also, since this system include a function to detect a period of voice signals, the robot can know when to start finding sounds direction and performing speech recognition automatically. In order to verify our systems performances, we install the proposed audition system to the prototype robot, called IROBAA (intelligent robot for active audition), and describe the experimental results.

Detection of multi-active markers and pose for formation control

This paper represents a part of outdoor localization method based on relative information: heading angle and distance to apply to formation control for multi-agent robots. Well known are the researches about formation control for multi-agent robots using various sensors: window GPSs, ultra-sonic sensors and a camera with color markers, but those methods mostly could be applied on special environment and have restrictions or much expensive. The proposed method would be expected to overcome those problems. To avoid electronic or natural illumination disturbance in the image-processing, we used active markers (four bright LED arrays) and considered only their ON/OFF time period to obtain the pose. In this paper, experimental results will show that the algorithm is possible to be applied.

Speaker localization among multi-faces in noisy environment by audio-visual integration

In this paper, we not only developed a reliable sound localization system including VAD (voice activity detection) component using three microphones but also a face tracking system using a vision camera. Moreover, we proposed a way to integrate these systems in the human-robot interaction to compensate the errors in the localization of a speaker and to reject unnecessary speech or noise signals entering from the undesired directions effectively. For the purpose of verifying our systems performances, we installed the proposed audition and vision system to the prototype robot, called IROBAA (Intelligent ROBot for Active Audition), and showed how to integrate an audio-visual system

Near minimum-time direct voltage control algorithms for wheeled mobile robots with current and voltage constraints

Near minimum-time direct voltage control (DVC) algorithms synthesizing path-planning and path-following are proposed for wheeled mobile robots (WMRs) satisfying (i) initial and final postures and velocities as well as (ii) voltage and current constraints. To overcome nonholonomic and nonlinear properties of WMRs, we divide our control algorithm for cornering motion into three sections: TSD (Translational Section of Deceleration), RS (Rotational Section), and TSA (Translational Section of Acceleration). We developed off-line DVC algorithms using the quadratic problem with the object function minimizing the total time, where voltages to the motors are controlled directly without velocity/torque-servo modules, while satisfying the current constraints. Two methods of searching for the two control parameters (number of steps for RS MR and velocity constraint in RS b) were considered: The one is composed of one simple 1-dimensional search for b, and the other is composed of two 1-dimensional searches for MR and b which has better performance. Performances of the proposed control algorithms are validated via various simulations.

Sound source localization in reverberant environment using visual information

Recently, many researchers have carried out works on audio-video integration. It is worth exploring because service robots are supposed to interact with human beings using both visual and auditory sensors. In this paper, we propose an audio-video method for sound source localization in reverberant environment. Using visual information from a vision camera, we could train our audio localizer to distinguish a real source from fake sources and improved the performance of audio localizer in reverberant environment.

A robot-assisted behavioral intervention system for children with autism spectrum disorders

The purpose of this paper is to propose and examine the feasibility of a robot-assisted intervention system capable of facilitating social training for children with autism spectrum disorder (ASD) via human-robot interaction (HRI) architecture. Based on the well-known discrete trial teaching (DTT) protocol for the therapy of children with ASD, our control architecture configures four modules-human perception, user input, the interaction manager, and the robot effector-such that the robot system generates differentiated training stimuli using motivation and Stroop paradigms and automatically copes with the childs response by using reliable human recognition and interaction technologies. Using these configurations, the proposed system performs the role of training the social skills of basic eye contact and reading emotions in children with ASD. By examining reliable performance evaluations and the positive effect of the training process targeting preschoolers with a high functioning level, we then verify that the proposed system can induce a positive transition in the response of children with ASD and the possibility of a labor-saving effect in carrying out autism treatments. Robot-assisted behavioral intervention based on automated interaction technologies is proposed.Socially validated therapeutic protocol is applied to child-robot interaction.Reliable eye contact detection is reviewed as clinical evidence.Differentiated robotic stimulation is designed to attract childrens attention.Reinforcement procedure with coping strategy is verified in repetitive training routines for clinical trials.

Speaker selection and tracking in a cluttered environment with audio and visual information

Presented in this paper is a data association method using audio and visual data which localizes targets in a cluttered environment and detects who is speaking to a robot. A particle filter is applied to efficiently select the optimal association between the target and the measurements. State variables are composed of target positions and speaking states. To update the speaking state, we first evaluate the incoming sound signal based on cross-correlation and then calculate a likelihood from the audio information. The visual measurement is used to find an optimal association between the target and the observed objects. The number of targets that the robot should interact with is updated from the existence probabilities and associations. Experimental data were collected beforehand and simulated on a computer to verify the performance of the proposed method applied to the speaker selection problem in a cluttered environment. The algorithm was also implemented in a robotic system to demonstrate reliable interactions between the robot and speaking targets.

Matching pursuit based robust acoustic event classification for surveillance systems

Acoustic event classification using matching pursuit and random forest.Event based feature extraction using atom time and frequency information.Superior in classification of human scream from other classes. Display Omitted The ability to automatically recognize acoustic events in real world conditions is an important application of the surveillance systems. This paper presents an acoustic event classification (AEC) method which uses the Matching Pursuit to extract the important Gabor atoms from input audio signals. Rather than extracting features in short-time frames, we apply the matching pursuit to the whole duration of an acoustic event. Information from atoms, such as time, frequency, and amplitude are used to construct time-frequency features. These features capture both spectral and temporal information of the sound event, which is analogous to the spectrogram representation. Experiments were performed on an 8-class database including human scream and gunshot. Under noisy and mismatched conditions, the proposed classification method achieves F1-score of 0.814, which is superior to state-of-the-art methods.