Yasuhiro Takishima

Reversible variable length codes

Proposes some reversible variable length codes (RVLCs) which can be decoded instantaneously both in the forward and backward directions and have high transmission efficiency. These codes can be used, for example, in the backward reconstruction of video signals from the data last received when some signal is lost midway in the transmission. Schemes for a symmetrical RVLC requiring only a single code table and for an asymmetrical RVLC having short average code length are introduced. They compare favorably with other reversible codes such as B2 codes in several aspects. >

Context-Aware Preference Model Based on a Study of Difference between Real and Supposed Situation Data

We propose a novel approach for constructing statistical preference models for context-aware recommender systems. To do so, one of the most important but difficult problems is acquiring sufficient training data in various contexts/situations. Particularly, some situations require a heavy workload to set them up or to collect subjects under those situations. To avoid this, often a large amount of data in a supposed situation is collected, i.e., a situation where the subject pretends/imagines that he/she is in a specific situation. Although there may be difference between the preference in the real situation and the supposed situation, this has not been considered in existing researches. Here, to study the difference, we collected a certain amount of corresponding data. We asked subjects the same question about preference both in the real and the supposed situation. Then we proposed a new model construction method using a difference model constructed from the correspondence data and showed the effectiveness through the experiments.

A study on rate distortion optimization scheme for JVT coder

This paper focuses on the encoder of ITU-T H.264 | ISO/IEC MPEG-4 Part-10 AVC (hereafter “JVT”), especially the Lagrange parameter for the rate distortion optimization (RDO) scheme. Since the Lagrange optimization method is very effective from the viewpoint of coding efficiency optimization, it is specified as an essential experimental condition for performance evaluation in JVT standardization activity. After discussing the background of this mechanism, we verify this method and the currently used parameter theoretically. Computer simulation results show that the Lagrange multiplier might not necessarily be original value; smaller values provide better performance in QCIF, for example.

Error states and synchronization recovery for variable length codes

An efficient technique for obtaining a variable length code (VLC) resistant to synchronization loss due to channel errors is proposed. Bit errors in the channel lead to incorrect decoding of a VLC and result in error states for continuous codewords. In this paper first a precise value is found for assessing the ability of a VLC to recover synchronization from the error states. Next a new transformation for converting a code into another of the same efficiency (equivalent transformation) is introduced, and several VLCs with an identical efficiency are compared through estimation of synchronization recovery in order to find an error-resistant VLC. Based on the equivalent transformation, an efficient method for finding an error-resistant VLC with high recovery performance from among multiple VLCs is described. >

Visual Cryptography Based on Void-And-Cluster Halftoning Technique

In this paper, visual cryptography for halftone images is proposed, which uses threshold arrays generated by a void-and-cluster method. The proposed halftone visual cryptography (HFC) using threshold arrays shows great flexibility for a combination of share images in order to decode a secret image. Moreover, it also achieves high quality share images, high speed processing and extensibility to (k,n) the visual cryptography of three share image categories, such as (2,3) or (3,3). Simulation results show that all the above features are achieved, while a secret image is clearly decoded, and the secret image cannot be seen on each share image.

Halftone Visual Cryptography Embedding a Natural Grayscale Image Based on Error Diffusion Technique

This paper proposes a novel visual cryptography method that embeds a natural secret image into two natural carrier images with high quality at low computational complexity. The secret image is decoded visually by superposing the two carrier images. For some applications such as entertainment use, all three images should be high quality natural images and low computational complexity is also required. However, in the conventional methods, picture quality and processing speed have been tradeoff relations. To solve this problem, we propose a method which is based on conventional visual cryptography using error diffusion halftoning and introduce an additional feedback mechanism to embed the high quality natural image. Experimental results show that the proposed method can achieve higher quality natural images of a superposed image and carrier images while computational complexity is maintained low when compared with the conventional methods.

Constructing a landmark identification system for Geo-tagged photographs based on Web data analysis

In this research, we propose a method to automatically generate a landmark identification system for geo-tagged photographs, based on analysis of various data collected from the Web. The method first conducts Web analysis based on three major procedures: (1) Automatic extraction of pointsof- interest (POIs) based on geographical clustering of geotagged images, (2) Retrieval of landmark candidates for each extracted POI from search results of map search API, and (3) Collection and feature extraction of Web images of the landmark candidates. The system then identifies the landmark which appears in the query geo-tagged photograph, by comparing the location and content-based features of the query with the information accumulated by the previous procedures. Experimental results show that the proposed method is capable to construct a highly accurate landmark identification system by leveraging Web information.

H.264 native video watermarking method

A novel video watermarking method is proposed, which can directly embed a watermark into a bit stream generated by an H.264 video encoder. Watermarking to encoded data provides a content provider of flexible capability to attach secret information; however it brings a drift problem. A drift compensation method has been proposed, however it requires many computations such as DCT and motion compensation. Therefore a new drift compensation method with spatiotemporal watermarking is proposed that only cancels the effect of the embedding on blocks with little motion. It achieves great reduction of computation and realizes fast and high picture quality watermarking. The computer simulation shows that degradation of picture quality is negligible compared to noises due to H.264 compression, and that the proposed method is highly resilient to frequency domain attacks

A fast video stitching method for motion-compensated frames in compressed video streams

We propose a fast video stitching method based on global motion tracking for motion-compensated frames, which can construct high definition or panorama video from standard definition video streams. Our method consists of two stages. The first stage calculates projection matrix between stitched frames coarsely by compensating for global motion of each input video stream. Since the result of the first stage may include some matching error, fine adjustment is performed in the second stage.

ATM networking and video-coding techniques for QOS control in B-ISDN

Integration of video services are addressed. The view that various types of end-to-end quality of service (QOS) for video communications, ranging from videophone to high-definition television (HDTV), should be supported to utilize broadband integrated services digital network (B-ISDN) services and video data is expressed. End-to-end QOS is recognized as a function of terminal capability and network performance. The key feature of B-ISDN is that both terminal capability and network performance are selectable and controllable. In order to capture the basic characteristics of end-to-end QOS, it is necessary to conduct experiments using a prototype system. A prototype system comprising an asynchronous transfer mode (ATM) switch and video codecs is presented. The ATM switch is based on distributed-memory-switch architecture and it will provide for nonblocking cell-switching capability with a very low cell-transfer delay. Developed codecs conform to H.261 with an additional function of controlling average coding bit rate. The coding bit rate is chosen as a parameter of terminal capability and the cell-loss ratio is chosen as a parameter of network performance. From experiments using a prototype system, a relationship between coding bit rate and cell-loss ratio is obtained. >