Eunjeong Ko

An Intelligent Wheelchair to enable mobility of severely disabled and elder people

In this paper, we develop an Intelligent Wheelchair (IW) system for the severely disabled people. The main function of our system is two-fold: 1) user intention recognition using vision technique and 2) sensor-based obstacle detection and path finding. The proposed system enable a user to control IW using his mouth shape and face movement. Furthermore, to fully guarantee users safety, the 10 range sensors are used to detect obstacles in environment and avoid them. To assess the effectiveness of the proposed IW, it was tested with 34 users and then the results show that it can provide a user unable to drive a standard joystick with friendly and convenient system.

Highway traffic flow prediction using support vector regression and Bayesian classifier

With the vast availability of traffic sensing data on highway, real-time traffic flow prediction is essential part of transportation, traffic control, reports of accidents and intelligent transportation systems. To satisfy the demand of traffic flow prediction, this paper presents the method of real-time traffic flow prediction based on Bayesian classifier and support vector regression (SVR). We first model the traffic flow and its relations on the roads using 3D Markov random fields in spatiotemporal domain. Based on their relations, we define cliques as combination of current road and its neighbors. The dependencies on the defined cliques are estimated by using multiple linear regression and SVR. Finally, the traffic flow at next time stamp is predicted by finding the speed level with decreasing the energy function. To evaluate the performance of the proposed method, it was tested on traffic data obtained from Gyeongbu expressway. The experimental results showed that the approach using SVR-based estimation had superior accuracy than linear-based regression.

Situation-based indoor wayfinding system for the visually impaired

This paper presents an indoor wayfinding system to help the visually impaired finding their way to a given destination in an unfamiliar environment. The main novelty is the use of the users situation as the basis for designing color codes to explain the environmental information and for developing the wayfinding system to detect and recognize such color codes. Actually, people would require different information according to their situations. Therefore, situation-based color codes are designed, including location-specific codes and guide codes. These color codes are affixed in certain locations to provide information to the visually impaired, and their location and meaning are then recognized using the proposed wayfinding system. Consisting of three steps, the proposed wayfinding system first recognizes the current situation using a vocabulary tree that is built on the shape properties of images taken of various situations. Next, it detects and recognizes the necessary codes according to the current situation, based on color and edge information. Finally, it provides the user with environmental information and their path through an auditory interface. To assess the validity of the proposed wayfinding system, we have conducted field test with four visually impaired, then the results showed that they can find the optimal path in real-time with an accuracy of 95%.

3D Markov Process for Traffic Flow Prediction in Real-Time

Recently, the correct estimation of traffic flow has begun to be considered an essential component in intelligent transportation systems. In this paper, a new statistical method to predict traffic flows using time series analyses and geometric correlations is proposed. The novelty of the proposed method is two-fold: (1) a 3D heat map is designed to describe the traffic conditions between roads, which can effectively represent the correlations between spatially- and temporally-adjacent traffic states; and (2) the relationship between the adjacent roads on the spatiotemporal domain is represented by cliques in MRF and the clique parameters are obtained by example-based learning. In order to assess the validity of the proposed method, it is tested using data from expressway traffic that are provided by the Korean Expressway Corporation, and the performance of the proposed method is compared with existing approaches. The results demonstrate that the proposed method can predict traffic conditions with an accuracy of 85%, and this accuracy can be improved further.

Predicting Spatiotemporal Traffic Flow Based on Support Vector Regression and Bayesian Classifier

Recently, with the rapid development of sensor technologies, it is important to manage the large amounts of traffic data and predict the traffic condition from them. To satisfy the demand of traffic flow estimation, this paper studies the method of real-time traffic flow prediction based on Bayesian classifier and support vector regression (SVR). We first model the traffic flow and its relations on the roads using 3D Markov random field in spatiotemporal domain. Based on their relations, we define cliques as combination of current cone-zone and its neighbors. The dependencies on the defined cliques are estimated by using multiple linear regression and SVR. Finally, the traffic flow at next time stamp is predicted by finding the speed level with decreasing the energy function. To evaluate the performance of the proposed method, it was tested on traffic data obtained from Gyeongbu expressway. The experimental results showed that the approach using SVR-based estimation showed superior accuracy than linear-based regression.

A Vision-Based Wayfinding System for Visually Impaired People Using Situation Awareness and Activity-Based Instructions

A significant challenge faced by visually impaired people is ‘wayfinding’, which is the ability to find one’s way to a destination in an unfamiliar environment. This study develops a novel wayfinding system for smartphones that can automatically recognize the situation and scene objects in real time. Through analyzing streaming images, the proposed system first classifies the current situation of a user in terms of their location. Next, based on the current situation, only the necessary context objects are found and interpreted using computer vision techniques. It estimates the motions of the user with two inertial sensors and records the trajectories of the user toward the destination, which are also used as a guide for the return route after reaching the destination. To efficiently convey the recognized results using an auditory interface, activity-based instructions are generated that guide the user in a series of movements along a route. To assess the effectiveness of the proposed system, experiments were conducted in several indoor environments: the sit in which the situation awareness accuracy was 90% and the object detection false alarm rate was 0.016. In addition, our field test results demonstrate that users can locate their paths with an accuracy of 97%.

Recognizing the sentiments of web images using hand-designed features

Recently, understanding sentiment expressed in social images and multimedia has attracted increasing attention by researchers. For sentiment analysis of social image, we should identify the visual features with high relations to human sentiments and then conduct analysis based on such visual features. Here, two visual vocabularies are built from color compositions and SIFT (scale-invariant feature transform) descriptors. Thereafter, the pLSA (probabilistic latent semantic analysis)-learning is employed to predict the human sentiment hidden in social images from visual words. The proposed system was evaluated to the images collected from Photo.net and Google and 15 Kobayashis sentiments were considered to label the images. The results were compared with man-labeled ground truth and then the proposed method shows the performance with an F1-measure results of above 70%.

Real-Time Highway Traffic Flow Estimation Based on 3D Markov Random Field

Nowadays, traffic flow estimation is the one of the most important topics in intelligent transportation systems (ITS). Accordingly, we propose a traffic flow estimation method using time-series analysis and geometric correlation. Firstly, we define a 3D heat-map to present the traffic state and spatial and temporal adjacent traffic condition. Thereafter, we model the dependency heat-map using spatiotemporal Markov Random Field and estimate the probability using logistic regression. To evaluate the performance of the proposed method, it was tested using data collected from expressway traffic that were provided by the Korean Expressway Corporation, and its performance was compared with those of other existing approaches. The results showed that the proposed method has a superior accuracy to others method, which has the accuracy of 85%.

Summarizing Social Image Search Results using Human Affects

In this paper, we propose the selection of representative images based on human affects. For this, the images are first transformed into the affective space using convolutional neural network (CNN). Thereafter, images are clustered on affective space and then the resulting clusters are ranked based on the proposed three properties ? coverage, affective coherence and distinctiveness. Finally, some representative images are selected from top-ranked clusters. The experiments conducted on Flickr images showed the effectiveness of the proposed method.

Discovering visual features for recognizing user's sentiments in social images

Recently, with the increasing of users and activities in social network service, an image sentiment analysis has been an important keyword for psychological study and commercial marketing. To recognize accurately users sentiments of the image, it is essential to identify discriminative visual features and then conduct analysis based on observed features. In this paper, we propose two hand-designed features: color composition and SIFT-based shape descriptor. These features are designed based on psychological study and experiments. First, two visual dictionaries are built by Kobayashis color image scale and Hierarchical clustering. Next, color compositions and SIFT-based descriptors are extracted from image. Then, the set of extracted features are separately transformed into a histogram representation by calculating the occurrences of the respective feature assigned to each visual word in the dictionary. To verify the effectiveness of the proposed features, we apply them to image sentiment analysis for predicting users polarity and affects. The recognition results were compared with man-labeled ground truth and then showed the performance with an F1-measure results of above 93%.