Haris Al Qodri Maarif

Malaysian Sign Language database for research

Malaysian Sign Language (MSL) is the main language that is commonly used by the hearing and speech impaired person in Malaysian. The SL (SL) involves hand movement, and hand gestures. In order to help people who are not familiar, but need to understand a particular SL, an automatic SL recognition system is highly required. The research in this area, especially for MSL, has been conducted by many researchers, but one of the main challenges in this research is the availability of suitable sign database for the recognition. The existing databases, especially which of MSL database, are provided often without a proper standard of image resolution, structure and compression that are sufficiently good for research purpose. To provide comprehensive information for the research on MSL, the MSL database is highly required. In this project, a MSL database is developed. The database is the first of the kind and developed for research purpose. In general, the structure of the MSL database is classified into groups that deal with the hand movement, hand gestures, and hand location. For the classification in our proposed database, the MSL is classified into One Hand, Two Hands, Static, and Dynamic. This classification is made to ease researchers in defining the research method for each type offhand signing.

Word classification for sign language synthesizer using hidden Markov model

Sign Language Synthesizer is an algorithm developed to provide signing animation from verbal/spoken language. Word classification in Natural Language Processing (NLP) is required to determine grammatically processed sentences for sign language synthesizer. The correct word position of output can provide understanding to users who use sign language synthesizer tools. In this paper, the Hidden Markov Model is proposed and implemented to process the words and locate their corresponding position correctly. The classification was done for Malay language and has resulted in an average accuracy of 74.67 %.

Complexity Algorithm Analysis for Edit Distance

Natural Language Processing (NLP) is a method which works on any language processing. Some of the algorithms are based on edit distance analysis. It is a process where the statistical calculations between two words or sentences are analyzed. Some of used edit distances for NLP are Levenshtein, Jaro Wrinkler, Soundex, N-grams, and Mahalanobis. The evaluation of edit distance is aimed to analyze the processing time of each edit distance in calculation of two different words or sentences. The objective of this paper is to evaluate the complexity of each distance, based on the time process.

Towards a Sign Language Synthesizer: a Bridge to Communication Gap of the Hearing/Speech Impaired Community

Sign language synthesizer is a method to visualize the sign language movement from the spoken language. The sign language (SL) is one of means used by HSI people to communicate to normal people. But, unfortunately the number of people, including the HSI people, who are familiar with sign language is very limited. These cause difficulties in the communication between the normal people and the HSI people. The sign language is not only hand movement but also the face expression. Those two elements have complimentary aspect each other. The hand movement will show the meaning of each signing and the face expression will show the emotion of a person. Generally, Sign language synthesizer will recognize the spoken language by using speech recognition, the grammatical process will involve context free grammar, and 3D synthesizer will take part by involving recorded avatar. This paper will analyze and compare the existing techniques of developing a sign language synthesizer, which leads to IIUM Sign Language Synthesizer.

Low Bit Rate Video Coding for Remote Desktop Operation of CNC Milling

Scalable Video Coding (SVC) is the extension of the H.264/AVC standard. The SVC has good capability in video transmission systems because of its scalability which can adapt in different network conditions, especially in low bit rate transmission. The objective of this paper is to develop the low bit rate video coding for assisting the remote operation of Command and Control (CNC) milling process through virtual simulation and remote desktop interface. The algorithm for low bit rate video coding will be in the downsampling of high resolution layer before the encoding process and upsampling the high layer data in the decoder. Based on simulation results, the proposed scheme shows good performance in maintaining the video quality at 128 kbps while providing high time savings for the encoding process.

Fast mode decision for scalable video coding over wireless network

The scalable video coding H.264/SVC is the video coding standard which is an extension of H.264/AVC. Because of its scalability, H.264/SVC has gained a great interest in video transmission. Partial bit stream can be transmitted and decoded by the H.264/SVC in order to provide quality in lower temporal or spatial resolutions or reduced fidelity while retaining a reconstruction quality that is high relative to the rate of the partial bit streams. Hence, H.264/SVC provides functionalities such as graceful degradation in lossy transmission environments, i.e. wireless networks, as well as bit rate format, and power adaptation. This paper deals with the Medium-Grain SNR Scalability (MGS) scheme and fast coding mode decision. The MGS scheme with two enhancement layers is applied for enhancing the streaming video quality. To perform an optimum mode decision, motion estimation is performed for all Macro Block (MB) modes, and the Rate Distortion (RD) costs are compared to identify an MB mode with the smallest RD cost. This increases computational complexity of H.264/SVC encoding. Therefore, fast mode decision algorithm scheme was implemented to speed up the encoding time. From the experiment, the fast mode decision was able to decrease the encoding process up to 30.22%.