We will describe interactive feature on an electronic holography system in this paper. In addition, we will introduce our developed 300-camera array in dense arrangement, the images captured by which are used to generate the holograms for the electronic holography system. Since the observer can specify viewpoint for observing the 3D objects using interactive feature, he or she could intuitively understand the 3D information more easily. This approach would have a variety of applications such as, for example, enabling a precious object in a museum to be viewed from desired direction.

This paper introduces two approaches for 3D visual systems toward ultra-realistic communication. The first system uses integral photography to capture ray information at slightly separated locations. The second system uses camera array that includes 300 cameras to capture ray information at more sparse locations than integral photography. Both systems use electronic holography as an ideal 3D display. Since spatial light modulators are currently inadequate specifications for electronic holography, the setup to remove disturbing light is necessary, which we implemented in both systems.

This paper introduces two 3D visual systems using ray-based image sensors and electronic holography display toward ultra-realistic communication. The first system uses integral photography for the capture of ray information at slightly separated locations. The second system uses camera array that includes 300 cameras to capture ray information at more sparse locations than integral photography. Both systems use electronic holography as an ideal 3D display. Since spatial light modulators are currently inadequate specifications for electronic holography, the setup to remove disturbing light is necessary, which we implemented in both systems.

An increasing number of projects in neuroscience require statistical analysis of high-dimensional data, as, for instance, in the prediction of behavior from neural firing or in the operation of artificial devices from brain recordings in brain-machine interfaces. Although prevalent, classical linear analysis techniques are often numerically fragile in high dimensions due to irrelevant, redundant, and noisy information. We developed a robust Bayesian linear regression algorithm that automatically detects relevant features and excludes irrelevant ones, all in a computationally efficient manner. In comparison with standard linear methods, the new Bayesian method regularizes against overfitting, is computationally efficient (unlike previously proposed variational linear regression methods, is suitable for data sets with large numbers of samples and a very high number of input dimensions) and is easy to use, thus demonstrating its potential as a drop-in replacement for other linear regression techniques. We evaluate our technique on synthetic data sets and on several neurophysiological data sets. For these neurophysiological data sets we address the question of whether EMG data collected from arm movements of monkeys can be faithfully reconstructed from neural activity in motor cortices. Results demonstrate the success of our newly developed method, in comparison with other approaches in the literature, and, from the neurophysiological point of view, confirms recent findings on the organization of the motor cortex. Finally, an incremental, real-time version of our algorithm demonstrates the suitability of our approach for real-time interfaces between brains and machines.

Neighboring views must be highly correlated in multiview video systems. We should therefore use various neighboring views to efficiently compress videos. There are many approaches to doing this. However, most of these treat pictures of other views in the same way as they treat pictures of the current view, i.e., pictures of other views are used as reference pictures (inter-view prediction). We introduce two approaches to improving compression efficiency in this paper. The first is by synthesizing pictures at a given time and a given position by using view interpolation and using them as reference pictures (view-interpolation prediction). In other words, we tried to compensate for geometry to obtain precise predictions. The second approach is to correct the luminance and chrominance of other views by using lookup tables to compensate for photoelectric variations in individual cameras. We implemented these ideas in H.264/AVC with inter-view prediction and confirmed that they worked well. The experimental results revealed that these ideas can reduce the number of generated bits by approximately 15% without loss of PSNR.

We propose a new multi-view video coding method using adaptive selection of motion/disparity compensation based on H.264/AVC. One of the key points of the proposed method is the use of view interpolation as a tool for disparity compensation by assigning reference picture indices to interpolated images. Experimental results show that significant gains can be obtained compared to the conventional approach that was often used

An increasing number of projects in neuroscience requires the sta- tistical analysis of high dimensional data sets, as, for instance, in predicting behavior from neural firing or in operating artificial de- vices from brain recordings in brain-machine interfaces. Linear analysis techniques remain prevalent in such cases, but classical linear regression approaches are often numerically too fragile in high dimensions. In this paper, we address the question of whether EMG data collected from arm movements of monkeys can be faith- fully reconstructed with linear approaches from neural activity in primary motor cortex (M1). To achieve robust data analysis, we develop a full Bayesian approach to linear regression that auto- matically detects and excludes irrelevant features in the data, reg- ularizing against overfitting. In comparison with ordinary least squares, stepwise regression, partial least squares, LASSO regres- sion and a brute force combinatorial search for the most predictive input features in the data, we demonstrate that the new Bayesian method offers a superior mixture of characteristics in terms of reg- ularization against overfitting, computational efficiency and ease of use, demonstrating its potential as a drop-in replacement for other linear regression techniques. As neuroscientific results, our anal- yses demonstrate that EMG data can be well predicted from M1 neurons, further opening the path for possible real-time interfaces between brains and machines.

Results of RNA secondary structure prediction algorithm are usually given as a set of hydrogen bonds between bases. However, we cannot know the precise structure of an RNA molecule by only knowing which bases form hydrogen bonds. One way to understand the structure of an RNA molecule is to visualize it using a planar graph so that we can easily know the geometric relations among the substructures such as stacking regions and loops. To do this, we consider bases to be particles on a plane and introduce a repulsive force and an attractive force among these particles and determine their positions according to these forces. A naive algorithm requires O(N2) time but we can reduce it to O(NlogN) with an approximation algorithm which is often used in the area of N-body simulation. Our program is written in parallel object-oriented language 'Schematic' which is recently developed. Efficiency of our implementation on a parallel computer and results of visualization of secondary structures are presented using cadang-cadang coconut viroid as an example.

An RNA secondary structure prediction method using a highly parallel computer is reported. We focus on finding thermodynamically stable structures of a single-stranded RNA molecule. Our approach is based on a parallel combinatorial method which calculates the free energy of a molecule as the sum of the free energies of all the physically possible hydrogen bonds. Our parallel algorithm finds many highly stable structures all at once, while most of the conventional prediction methods find only the most stable structure. The important idea in our algorithm is search tree pruning, with dynamic load balancing across the processor elements in a parallel computer. Software tools for visualization and classification of secondary structures are also presented using the sequence of cadang-cadang coconut viroid as an example. Our software system runs on CM-5.

We have defined the folding index of the RNA secondary structure and compared the value of this index obtained from a given RNA molecule with one from the random sequences with the same A, U, G, C ratio. The mRNA of the cytochrome c oxidase subunit II and cytochrome P450 from various species was analyzed. We found that the mRNA of these genes were more folded than the random sequences and the coding regions of these mRNA were as equally folded as the random sequences. This result suggests that the coding region of these mRNA has no more important secondary structure than have the random sequences with the same A, U, G, C ratio.

We present a new algorithm for the display of RNA secondary structure. The principle of the algorithm is entirely different from those currently in use in that our algorithm is ‘object oriented’ while currrent algorithms are ‘procedural’. The circular RNA molecule of chrysanthemum stunt viroid was used as input data for demonstrating the operation of the program. The major interest of this method will be found in its potential use in simulation graphics of RNA folding processes.

We have previously reported an algorithm (Yamamoto and Yoshikura, 1985) for the prediction of optimum and suboptimum RNA folding structures. The calculation data was presented as an 'information map'. However, the result was affected by the starting point of calculation. In this paper, we have improved the method so that the result will not be affected by the starting point. In addition, we present a method of converting the information map into a set of pictures of optimal and suboptimal molecular structures.

A vector diagram was devised for representation of protein structure. This was done by giving each nucleic acid triplet a corresponding vector, and expressing a protein as a set of vectors which were aligned in the order of peptide sequence. The magnitude of each vector was made to reflect the hydrophilicity of amino acid, and the direction of the vector was chosen so that similar amino acids had similar directions. The diagram not only shows the primary sequence but suggests the folding structure of proteins.

We present an algorithm for prediction of RNA secondary structures. The program consists of three parts: the first computes location and free energy of every possible stem-loop structure, the second computes probability of its formation, and the third lists the positions and free energies of all the stem-loops in the order of their probability sizes. The circular RNA molecule of chrysanthemum stunt viroid was used as an input data for demonstrating the operation of the program.