http://dspace.cityu.edu.hk:80/handle/2031/740 Fri, 31 May 2013 18:28:09 GMT 2013-05-31T18:28:09Z http://dspace.cityu.edu.hk:80/handle/2031/4729 Title: The development of an automatic real-time lipreading system Authors: Wang, Shilin (王士林) Abstract: Using visual information in automatic speech recognition has aroused the interest of many researchers in recent years because the visual information will help enhancing the robustness of the system. This thesis presents the studies of developing a real-time system for automatic speech recognition solely based on the visual cues from the lip shape and movement. Lip segmentation, modeling, visual feature extraction and recognition are the major issues of the system and new algorithms for these issues are presented in this thesis. For lip segmentation, most of the widely used methods are based on the color or intensity information. However, for the lip images with low contrast, these methods will not achieve satisfactory results. In our study, both the color information and spatial location are integrated into a fuzzy clustering framework. The lip image is represented in CIELAB and CIELUV color space, where the luminance and chromaticity information is separated and the distance between any two points in the color space is proportional to their perceptual color difference. By integrating the spatial information, the proposed algorithm can differentiate the pixels with similar color but located in different regions. From the experimental results, the proposed algorithm outperforms other lip segmentation techniques especially for images with low color contrast. An extension of the algorithm has also been developed to solve the lip segmentation problem with the presence of beards, which is regarded as a difficult problem for lip region segmentation. For the lip contour modeling and extraction, accuracy, robustness and efficiency are the primary concerns. A 16-point model is employed to describe the lip contour. Some geometric constraints are applied to ensure that the extracted lip contour is physically meaningful. Based on the membership distribution derived from the lip segmentation procedure, a region-based cost function is defined, which is much more robust than the edge-based and intensity-based cost functions. A point-driven optimization procedure with some fast implementation techniques is used for model fitting and thus the lip contour can be obtained in an efficient manner. A visual feature set containing the geometric parameters, lip shape descriptors, and inner mouth information is obtained from the lip model for visual speech recognition purpose. A spline representation is employed to translate the discrete- sampled visual features into the continuous domain. The spline coefficients in the same word class are constrained to have the same mean and covariance matrix and can be estimated from the training data by the EM algorithm. In the speaker independent recognition task, a multi-model approach is proposed to overcome the difficulty due to the large variation caused by different speakers. By comparing with the HMM, the proposed method gives better result especially when only limited training data is available. An automatic lipreading system has been implemented and running on a 1.9 GHz PC. An accuracy of 96% for the speaker dependent recognition and 88% for the speaker independent recognition have been achieved. With the efficient implementation of all the algorithms, the system is able to process images at a rate higher than 25 frames/sec, leaving room for additional tasks in real-time applications. Notes: xi, 134 leaves : ill. (some col.) ; 30 cm.; Thesis (Ph.D.)--City University of Hong Kong, 2004; Includes bibliographical references (leaves 120-127); CityU Call Number: TK7895.S65 W36 2004 Thu, 01 Jan 2004 00:00:00 GMT http://dspace.cityu.edu.hk:80/handle/2031/4729 2004-01-01T00:00:00Z http://dspace.cityu.edu.hk:80/handle/2031/4356 Title: Algorithmic approaches for IP address lookup and packet classification Authors: Liu, Cutson Ngai-hang (廖毅恒) Abstract: The basic function of an Internet router is to forward packets towards their destinations. As the applications in the Internet are evolving towards multimedia-oriented and QoS-aware, more complex analysis of the incoming packet is performed by the routers in addition to the IP address lookup. Routers are required to classify packets based on multiple header fields into equivalence classes called flows. The challenges for the IP address lookup and packet classification are driven by the following factors: (i) the rapid growth of communication bandwidth, (ii) the expected growth of the routing table and filter database, (iii) the deployment of IPv6, and (iv) fast incremental update. This thesis presents algorithmic approaches to solve these problems. The first contribution of this work is the design of an efficient hardware solution for IP address lookup. The problem is modeled as a searching problem on a binary-trie. The binary-trie is partitioned into fixed size non-overlapping sub-trees. Each sub-tree is represented using a bit-vector and can be searched in parallel for the best matching prefix. The address lookup is implemented using hardware pipeline with a throughput of one lookup per memory access. A distinguishing feature of our design is that routing table entries are not replicated in the data structure. Hence, table update can be done in constant time with only a few memory accesses. Our design can be extended to IPv6. The second contribution of this work is the design of a two-stage approach for packet classification. The first stage determines the best matching source and destination prefix pair and the second stage determines the highest priority matching filter by comparing the remaining fields against a short list of candidate rules in parallel. The first stage 2- dimenional search problem is reduced to a l-dimenional search problem by a process called filter decomposition. The decomposed filters are organized as a height-balanced search tree. The search operation is speeded up by parallel processing techniques to achieve a throughput of one packet per memory cycle. Our design is scalable to larger rule databases and IPv6. It also allows the incremental updates the data structures. Notes: CityU Call Number: TK5105.875.I57 L5847 2005; Includes bibliographical references (leaves 86-90); Thesis (Ph.D.)--City University of Hong Kong, 2005; ix, 90 leaves : ill. ; 30 cm. Sat, 01 Jan 2005 00:00:00 GMT http://dspace.cityu.edu.hk:80/handle/2031/4356 2005-01-01T00:00:00Z