Browsing by Author "Chang, Fuh-Yu"

Now showing 1 - 3 of 3

A fast learning algorithm for robotic emotion recognition
(2007)
Hong, Jung-Wei
;
Han, Meng-Ju
;
Song, Kai-Tai
;
Chang, Fuh-Yu
;
電控工程研究所
The capability of robotic emotion recognition is an important factor for human-robot interaction. In order to facilitate a robot to function in daily live environments, a emotion recognition system needs to accommodate itself to various persons. In this paper, an emotion recognition system that can adapt to new facial data is proposed. The main idea of the proposed learning algorithm is to adjust parameters of SVM hyperplane for learning emotional expressions of a new face. After mapping the input space to Gaussian-kernel space, support vector pursuit learning (SVPL) is applied to retrain the hyperplane in the new feature space. To expedite the retraining procedure, only samples classified incorrectly in previous iteration are combined with critical historical sets to restrain a new SVM classifier. After adjusting hyperplane parameters, the new classifier will recognize previous erroneous facial data. Experimental results show that the proposed system recognize new facial data with high correction rates after fast retraining the hyperplane. Moreover, the proposed method also keeps satisfactory recognition rate of old facial samples.
A new information fusion method for SVM-Based robotic audio-visual emotion recognition
(2007)
Han, Meng-Ju
;
Hsu, Jing-Huai
;
Song, Kai-Tai
;
Chang, Fuh-Yu
;
電控工程研究所
Emotion recognition has become an important research area for advanced human-robot interaction. Through recognizing facial expressions, a robot can interact with a person with a more friendly manner. In this paper, we proposed a bimodal emotion recognition system by combining image and speech information. A novel information fusion strategy is proposed to set proper weights to two feature modalities based on their recognition reliability. The fusion weights are determined by the distance between test data and the classification hyperplane and the standard deviation of training samples. After normalization using the mean distance between training samples and the hyperplane, the fusion weight is set to represent the classification reliability of individual modality. In the latter bimodal SVM classification, the recognition result with higher weight is selected. The complete procedure has been implemented in a DSP-based system to recognize five facial expressions on-line in real time. The experimental results show that a recognition rate of 86.9% is achieved, an improvement of 5% compared with using only image information.
Fabrication of an antireflective polymer optical film with subwavelength structures using a roll-to-roll micro-replication process
(2008-07-01)
Ting, Chia-Jen
;
Chang, Fuh-Yu
;
Chen, Chi-Feng
;
Chou, C. P.
;
機械工程學系
An antireflective optical film with subwavelength structures replicated by use of a roll-to-roll micro-replication process (RMRP) is investigated. Firstly, a single layer of a nanostructure on a polymer film is designed for an antireflection purpose by the finite difference time domain method in the visible light spectrum. Structures of a conical cylinder array, with spatial period of 400 nm, diameter of 200 nm and height of 350 nm, are numerically obtained. Then, such structures are fabricated by RMRP combining originated structure fabrication realized by deep ultra-violet lithography and dry etching, Ni mold electroplating and replication by using the roll-to-roll process imprinting into the flexible polyethylene terephthalate substrate. A nanostructure roller mold bonded with Ni molds has been successfully fabricated and coated with the self-assembly monolayer process for the purpose of fabricating an anti-adhesion film and improving the lifetime of the Ni molds. The duplicated nanostructure films show a good optical quality of antireflection (AR <= 2.45% in a 400-700 nm spectral range) and are in good agreement with the theoretical predictions. The experimental results show that the developed process is a promising and cost-effective method for the continuous duplication of flexible devices with nano-scaled feature sizes used in nanophotonics by RMRP.