Aishah Abdul Razak

Comparison between fuzzy and NN method for speech emotion recognition

This paper discusses an approach towards automatic recognition of emotion in speech which is adopted into a system named voice driven emotion recognizer mobile phone (VDERM). First, a design for the emotion recognizer is proposed, LPC analysis algorithm has been used for the speech emotion feature extraction. A total of 18 speech features have been selected to represent each emotion. A database consisting of emotional Malay and English, male and female voice samples have been developed for training and recognition purposes. Two recognition methods namely neural network and fuzzy model have been experimented and compared. The results show that both methods have their own advantage and disadvantage in application to emotion recognition. A recognition rate of up 60% is achievable by using these computer methods which is sufficient based on the recognition rate achieved by human.

Emotion pitch variation analysis in Malay and English voice samples

This paper discusses the pitch variation in Malay and English emotional voice samples for six emotion states. LP analysis is carried out to calculate the emotion features in speech and the results are observed in terms of average pitch, pitch range and jitter. Comparison between male and female pitch is also done. Based on the pitch variation analysis, language factor does not affect the acoustic correlates of emotional speech.

3D Modelling of a Famosa Fortress, Malaysia Based on Comparison of Textual and Visual Data

This paper presents an attempt to model the “A Famosa Fortress” in Malaysia into 3D. This building was built in 1511 by the Portuguese and went through several architectural developments and changes before being largely destroyed during the British occupation in 1824. The biggest challenge in this research is to determine the original fortress layout due to the lack of any authoritative documentation pertaining to the fortress. Detail analysis has been conducted to identify reliable sources for references which are available in the form of text and visual. In this paper, we focus on comparison of selected textual and visual data to come out with a verifiable conjectural layout of the fortress. We then pre-visualized the layout in 3D model. Some samples of the model are presented here however there are still rooms for improvements before it is finalized. The output of this research will be tested for application in tourism and education.

A preliminary speech analysis for recognizing emotion

Some speech analysis to extract emotion from voice is discussed. An emotional Malay and English voice database has been developed, consisting six basic emotions namely happiness, sadness, disgust, fear, anger and surprise. As the target is content independent emotion recognition, 4 short sentences that have the most natural meaning is adopted for the illustration and analysis. A study on speech prosody is done to identify the emotional features of voice. Variation on the samples energy, duration, and pitch for different emotions is compared. Spectrogram analysis is done on some samples to observe the effect of formant. It is found that duration, average energy and pitch can provide some indication of emotional content of a speech, but it is not enough to correctly represent the emotions. Even though there are slightly different pattern for English and Malay samples, it is still reasonable to assume that there are standard acoustic configurations in expressing particular emotions.

3D modelling of the A Famosa fortress in Melaka, Malaysia

This paper presented our approach to developing the 3d model of A Famosa Faortress in Melaka, Malaysia. As the physical remnant of the fortress are minimal, a traditional 3D modelling technique based on collected data has been adopted. This method has been adopted based on case studies of three comparable reconstruction projects namely Michelsberg synagogue, The Jewish Quarter in Regensburg and Reconstruction of the old main church in Curitiba, Brazil. In this paper we discussed the 3D modeling technique used and presented some output of the model.

Digital preservation of a famosa fortress in malaysia

In this paper, we described an attempt to preserve the A Famosa Fortress in Malaysia in 3D model. First, a brief background of the A Famosa fortress is presented to justify the importance of preserving this historical heritage. Then the stages involved in the digital preservation process are discussed and some images of the 3D model of this fortress are illustrated. To evaluate the importance of the digital preservation of A Famosa and to find out its potential application in the tourism and education industry in Malaysia, a human evaluation has been conducted and the results are presented and discussed in detail.

A study on practical approach of using motion capture and keyframe animation techniques

Producing 3D character animation requires proper planning and the right tools; these include highly technical manpower and facilities. Our paper discusses a practical approach for producing 3D character animation namely motion capture (Mocap) and keyframe animation. This research is based on our experience utilising both multimedia applications at Faculty of Creative Multimedia, Multimedia University. We analyse the strengths and weaknesses of these two methods based on a set of experiments, followed by conclusions that serve as a set of guidelines for animators as they decide on the use of motion capture or keyframe.

Voice Driven Emotion Recognizer Mobile Phone: Proposal and Evaluations

This article proposes an application of emotion recognizer system in telecommunications entitled voice driven emotion recognizer mobile phone (VDERM). The design implements a voice-to-image conversion scheme through a voice-to-image converter that extracts emotion features in the voice, recognizes them, and selects the corresponding facial expression images from image bank. Since it only requires audio transmission, it can support video communication at a much lower bit rate than the conventional videophone. The first prototype of VDERM system has been implemented into a personal computer. The coder, voice-to-image converter, image database, and system interface are preinstalled in the personal computer. In this article, we present and discuss some evaluations that have been conducted in supporting this proposed prototype. The results have shown that both voice and image are important for people to correctly recognize emotion in telecommunications and the proposed solution can provide an alternative to videophone systems. The future works list some modifications that can be done to the proposed prototype in order to make it more practical for mobile applications.

Adaptation of moveable pop-up technique in 3D animation

To convey a story or message, contemporary 3D animation utilizes a conventional continuity approach in which events are arranged in a smooth, sequential, and uninterrupted manner. The results of good continuity are coherence in visual perception and a seamless flow of action and events from one shot or sequence to another in all motion media. In making a film, continuity coupled with good editing principles and appropriate use of transitions in the post-production phase make for an organized shooting plan. In this article we discuss the potential of adapting a movable pop-up technique as a continuity element for the entire flow in 3D animation. This is done by integrating the technique within the animated story, transforming scenes, and narrating the story from one shot to another continuously by substituting basic editing transitions in a film. These findings demonstrate that movable pop-up animation affects animation timing and is incapable of achieving classical continuity (defined as making transitions as transparent as possible to the audience). Nevertheless, this technique is sufficient to create coherence in visual perception and establish a continuous flow of events within a distinct animation style.