Automatic text summarization has the goal of reducing the size of a document while preserving its content. Generally, producing a summary as extracts is achieved by including only sentences which are the most topic-related. DOCUSUM is our summarization system based on a new topic keyword identification method. The process of DOCUSUM is as follows. First, DOCUSUM converts the content words of a document into elements of a context vector space. It then constructs lexical clusters from the context vector space and identifies core clusters. Next, it selects topic keywords from the core clusters. Finally, it generates a summary of the document using the topic keywords. In the experiments on various compression ratios (the compression of 30%, the compression of 10%, and the extraction of the fixed number of sentences: 4 or 8 sentences), DOCUSUM showed better performance than other methods.

Analog Hopfield neural networks (HNNs) have so far been used to solve many kinds of optimization problems, in particular, combinatorial problems such as the TSP, which can be described by an objective function and some equality constraints. When we solve a minimization problem with equality constraints by using HNNs, however, the constraints are satisfied only approximately. In this paper we propose a circuit which rigorously realizes the equality constraints and whose energy function corresponds to the prescribed objective function. We use the SPICE program to solve circuit equations corresponding to the above circuits. The proposed method is applied to several kinds of optimization problems and the results are very satisfactory.

We have developed an ultrahigh-sensitivity "New Super-HARP" handheld camera, which has a sensitivity that is about 100 times as great as that of a CCD camera. The sensitivity of TV cameras is determined by the performance of the imaging device. We developed the world's first imaging device that achieves high sensitivity and high picture quality by using the avalanche multiplication phenomenon in an amorphous selenium photoconductive target. This "Super-HARP" pickup tube, which has already been used in TV production, has a selenium target 8-µm thick. It is about 10 times as sensitive as CCDs. We have now developed a greatly improved version of the Super-HARP tube with a target 25-µm thick. This improved version, called the New Super-HARP pickup tube, is about 10 times as sensitive as the Super-HARP pickup tube. The New Super-HARP handheld camera equipped with the new tubes has a maximum sensitivity of 11 lx at F8. This camera is a powerful tool for reporting breaking news at night and other low-light conditions, the production of scientific programs, and numerous other applications.

Several lifting implementation techniques for invertible deniterlacing are proposed in this paper. Firstly, the invertible deinterlacing is reviewed, and an efficient implementation is presented. Next, two deinterlacer-embedded lifting architectures of discrete wavelet transforms (DWT) is proposed. Performances are compared among several architectures of deinterlacing with DWT. The performance evaluation includes dual-multiplier and single-multiplier architectures. The number of equivalent gates shows that the deinterlacing-embedded architectures require less resources than the separate implementaion. Our experimental evaluation of the dual-multiplier architecture results in 0.8% increase in the gate count, whereas the separate implementation of deinterlacing and DWT requires 6.1% increase from the normal DWT architecture. For the proposed single-multiplier architecture, the gate count is shown to result in 4.5% increase, while the separate counterpart yields 10.7% increase.

This paper presents a framework for modeling and mixed-mode simulation of circuits/interconnects and electromagnetic (EM-) radiations. The proposed framework investigates the signal integrity in VLSI chips, packages and wiring boards at the GHz-band level, and verifies the electromagnetic interference (EMI) and the electromagnetic compatibility (EMC) of high-speed systems. In our framework, the frequency characteristics of interconnects and EM-radiations are extracted by the full-wave FDTD simulation. The macromodels of interconnects are synthesized as SPICE subcircuits, and the impulse responses of EM-radiations are stored in the database. Once the macromodels are synthesized, the circuit simulation with the consideration of EM-effects can be performed by using SPICE. The EM-field distributions can be also easily calculated by taking convolutions of pre-simulated EM impulse responses and the SPICE results.

In this paper, two types of multi-stage partial parallel interference cancellation (PIC) receivers are considered for multi-rate DS-CDMA system: multi-stage PIC receiver with partial cancellation factors and multi-stage PIC receiver with decision thresholds. Bit error rate (BER) of the multi-stage partial PIC receivers is obtained by simulation in a Rayleigh fading channel. It is shown that the multi-stage partial PIC receivers achieve smaller BER than the matched filter (MF) receiver, multi-stage PIC receiver, group-wise successive interference cancellation (GSIC) receiver, and extended GSIC receiver (EGSIC) for the multi-rate DS-CDMA system in a Rayleigh fading channel.

We have recently developed a programmable composite noise generator (P-CNG) which can easily control noise parameters such as average power, time-based amplitude probability distribution (APD), crossing rate distribution, occurrence frequency distribution and burst duration. Two applications of the P-CNG are demonstrated to show its usefulness. For the first application, Middleton's Class A noise is simulated. A method of setting parameters for Class A noise is demonstrated. The APD of P-CNG output is in good agreement with that of true Class A noise. In the second application, the P-CNG is used for subjective evaluation test (opinion test) of TV picture degradation. Five simple composite noise models with two kinds of APD are used. Other parameters such as average power are kept constant. Experimental results show that the envelope and APD of composite noises do not greatly influence the subjective evaluation. Finally the capabilities of the P-CNG are shown.

In this paper, we propose topic extraction models based on statistical relevance scores between topic words and words in articles, and report results obtained in topic extraction experiments using continuous speech recognition for Japanese broadcast news utterances. We attempt to represent a topic of news speech using a combination of multiple topic words, which are important words in the news article or words relevant to the news. We assume a topic of news is represented by a combination of words. We statistically model mapping from words in an article to topic words. Using the mapping, the topic extraction model can extract topic words even if they do not appear in the article. We train a topic extraction model capable of computing the degree of relevance between a topic word and a word in an article by using newspaper text covering a five-year period. The degree of relevance between those words is calculated based on measures such as mutual information or the χ2-method. In experiments extracting five topic words using a χ2-based model, we achieve 72% precision and 12% recall for speech recognition results. Speech recognition results generally include a number of recognition errors, which degrades topic extraction performance. To avoid this, we employ N-best candidates and likelihood given by acoustic and language models. In experiments, we find that extracting five topic words using N-best candidate and likelihood values achieves significantly improved precision.

The advances in services using the present Internet mean that there will be increasing demand for Video on Demand (VoD) on the Internet in the near future. However, because of the best-effort characteristics of the Internet, it is important to suppress the degradation of communication quality caused by packet dropping when Internet traffic is congested. This paper focuses on MPEG transmission over the Internet, and suitable control mechanisms are established for an acceptable Quality of Service (QoS) improvement through detailed evaluation. Packets are classified using a frame-based scheme. The server applies the proposed End-to-End control scheme and shuffles the order of packets to avoid burst dropping, and may omit selected packets belonging to certain frames prior to forwarding. At the intermediate routers, transferred packets are transmitted according to Round Robin (RR) or Weighted Round Robin (WRR) scheduling, and are dropped statistically using selective Random Early Detection (RED) corresponding to frame attributes when there is congestion. We evaluate the proposed performance of transmission method using both computer simulations and empirical measurements of picture quality. The results show that when the traffic volume cannot be estimated in the intermediate routers, the combined use of RR, shuffling and conditional RED is effective, and when the traffic volume can be estimated, the combination of WRR, rate control and RED is effective.

Gallager has defined an ensemble of regular low density parity check (LDPC) codes for deriving the ensemble performance of regular LDPC codes. The ensemble is called the Gallager ensemble. In this paper, we define a new ensemble of LDPC codes, called extended Gallager ensemble, which is a natural extension of the Gallager ensemble. It is shown that an extended Gallager ensemble has potential to achieve larger typical minimum distance ratio than that of the original Gallager ensemble. In particular, the extended Gallager ensembles based on the Hamming and extended Hamming codes have typical minimum distance ratio which is very close to the asymptotic Gilbert-Varshamov bound. Furthermore, decoding performance of an instance of an extended Gallager ensemble, called an extended LDPC code, has been examined by simulation. The results show good block error performance of extended LDPC codes.

Optical disk systems have been used in a wide range of applications and the performance of these systems has been improving rapidly. The optical integration is one of the important technologies of the progress of the optical pickup head in the system. It can make the optical pickup head miniature, light sized, ease to assemble, cost reduced and reliable during long time operation. In this paper, at first, merits and features of the optical integration for optical disk systems are briefly reviewed. Then, our activities on the development of the various hybrid-integrated optical devices, "Laser Couplers" are reported. Especially, the most recent results on the "Two-wavelength-beam Laser Coupler," are described in detail. It has two-wavelength laser emission/detection functions and easily realizes an ultimate solution of the optical pickup head for the DVD/CD system, now being used in "PlayStaion2. " To simplify the structure and the fabrication of this device, a new monolithic-integrated two-wavelength laser diode, called a visible and infrared laser diode, has been developed, which can be easily fabricated using only two steps of metal organic chemical vapor deposition. The structures and characteristics of this device are reported. Lastly, other recent technologies of the optical integration, including monolithically integration, near field optics and so on, are reviewed.

The economic success of monolithic optoelectronic and photonic ICs (OEICs, PICs) on InP depends strongly on their cost and performance. Taking into account these requirements as well as today's advanced hybrid integration approaches, the current status of monolithic InP integration technology is reviewed in examples, especially on components suitable for future telecommunication systems.

We report the measurements on the ramp-edge type Josephson and quasiparticle tunnel junctions with the different interface angle geometry using high-Tc YBa2Cu3O7-y (YBCO) electrodes. The YBCO/I/Ag tunnel junctions with different crystal-interface boundary angles are fabricated for the investigation of zero bias conductance peak. The angle dependent zero bias conductance peak typical to a dx2-y2-wave superconductor is observable. For Josephson junctions, YBCO ramp-edge junctions with different ab-plane electrodes relatively rotated by 45are fabricated using a CeO2 seed-layer technique. The temperature dependence of the maximum Josephson current for YBCO/PBCO/YBCO junctions (PBCO: PrBa2Cu3O7-y) exhibits angle-dependent behavior, qualitatively different from the Ambegaokar-Baratoff prediction. Under microwave irradiation of 9 GHz, the Shapiro steps appear at integer and/or half integer multiples of the voltage satisfying Josephson voltage-frequency relation, whose behavior depends on the sample angle geometry. The results are reasonably interpreted by the dx2-y2-wave theory by taking the zero energy state into account.

An optical detection system using a DFB laser with a very small aperture is theoretically proposed. The threshold gain level in DFB laser is sensitively varied with combined reflections by the facet and the corrugation as well as with the optical injection reflected at the surface of the optical disk. Variation of the threshold gain level is counted as the voltage change on electrodes of the laser. It is found that sensitivity of the optical detection with a well-designed DFB laser becomes six times larger than that with conventional Fabry-Perot ones. Field distribution around the small aperture is analyzed taking into account both the near-field and the radiation field. Numerical data on the voltage change are given as examples of the detection system.

This letter proposes a 3-D stereo endoscopic image processing system. Whereas a conventional 3-D stereo endoscopic system has simple monitoring functions, the proposed system gives doctors exact depth feelings by providing them depth value information, visualization, and stereo PACS viewer to aid an education, accurate diagnosis, a surgical operation, and to further apply in a robotic surgery.

In order to design oscillators and switches phase noise characteristic is the key to obtain high quality frequency spectrums. Since the phase noise is directly affected by the 1/f noise of transistors in the circuit, 1/f noise measurement and modeling are important. This paper describes 1/f noise measurement, frequency and bias dependent flicker noise model, and noise parameter extraction method of MOSFET's. Also, for MOSFET's geometry dependencies of drain current 1/f noise are analyzed and modeled. The model has been verified by measuring the noise current spectral density of MOSFET's in two different process devices.

Finding DC operating-points of nonlinear circuits is an important and difficult task. The Newton-Raphson method employed in the SPICE-like simulators often fails to converge to a solution. To overcome this convergence problem, homotopy methods have been studied from various viewpoints. The fixed-point homotopy method is one of the excellent methods. However, from the viewpoint of implementation, it is important to study it further so that the method can be easily and widely used by many circuit designers. This paper presents a practical method to implement the fixed-point homotopy method. A special circuit called the solution-tracing circuit for the fixed-point homotopy method is proposed. By using this circuit, the solution curves of homotopy equations can be traced by performing the SPICE transient analysis. Therefore, no modification to the existing programs is necessary. Moreover, it is proved that the proposed method is globally convergent. Numerical examples show that the proposed technique is effective and can be easily implemented. By the proposed technique, many SPICE users can easily implement the fixed-point homotopy method.

A novel inverse scattering approach is developed to the reconstruction of electrical property distributions of a two-dimensional biaxial anisotropic object using time-domain scattering data. The approach is an extension of the forward-backward time-stepping (FBTS) algorithm previously described for an isotropic object. Synthetic examples of inversion are given to assess the effectiveness of the proposed method.

Footprint assembly was proposed to reduce the blurriness of texture mapped image by mipmapping. Even though it can improve the quality of texture mapped image, there are yet blurring due to the limitation of it's filter kernel. This paper proposes a novel texture filtering, called adaptive footprint assembly (AFA), to overcome the limitation of footprint assembly. The proposed method greatly improves the quality of texture mapped images.

A low-profile bi-directional cavity antenna has been developed for the IMT-2000 indoor base stations. The geometrical relationships required for the design of an antenna with broadband impedance characteristics, which are obtained as a superposition of two resonant modes (M-antenna + metal case), are presented. The approximate equations describing the resonant frequencies associated with the two resonant modes are derived. By using the equations, a cavity antenna with dimensions of 120 mm 120 mm 12 mm and a fractional bandwidth of 18.3% (VSWR <2) that meets the IMT-2000 specification can be designed successfully. The proposed design procedure of the antenna is confirmed by the measurements.