The blendshape is an effective tool in computer facial animation, enabling represention of muscle actions. Limitations exist, however, in the level of realism attainable under conventional use of blendshapes as non-intersecting deformations. Using the principle of superposition, it is possible to create a facial model with overlapping blendshapes and achieve more realistic performance. When blendshapes overlap, the region of intersection is in superposition and usually exhibits undesired surface interference. In such cases we use a corrective blendshape to remove the interference automatically. The result is an animatable facial model implemented in Maya which represents the effects of muscle action superposition. Performance created with our model of a known human subject is compared to 3D scan reference data and video reference data of that person. Test animation is compared to video reference footage. The test animation seems to mimic the effects of actual muscle action superposition accurately.

No description available.

The research presented in this thesis represents an effort to characterize the properties of ultra-high molecular weight polyethylene (UHMWPE). As a composite of polymers, the properties of UHMWPE are time-dependent. It is desired by research sponsors to know the properties of the material at high strain rates, in order to simulate the use of these materials in computer models. Properties believed to be significant which are investigated in this research are the tensile properties of lamina and laminates, and the interlaminar shear properties of laminates. The efficacy of using time-temperature superposition to shift tensile properties of the composite is investigated, and a novel apparent shear strength test is proposed and demonstrated. The effects of processing the material at various temperatures and pressures are also investigated. / Master of Science

Thesis--Ph. D., University of Hong Kong, 1983.

published_or_final_version / Physics / Doctoral / Doctor of Philosophy

International Telemetering Conference Proceedings / October 22-25, 1984 / Riviera Hotel, Las Vegas, Nevada / For maritime applications of distress communications via geostationary satellites a special method of signal processing was developed, called superposition technique. The data frame containing the alert message is transmitted from a distress equipment repeatedly. After being relayed by the satellite the signal is detected and improved by superimposing the frames. Around 14 dB is the actual processing gain. Thus a distress buoy is able to transfer a message from all over the world with high reliability by only transmitting a power of 50 mW omnidirectionally over a slant range of about 40 000 km. The described system, called the Distress Radio Call System (DRCS) was tested in a Coordinated Trials Program (CTP) of 6 nations. Both in a simulation phase and in a field test under exactly the same environmental conditions the DRCS with its superposition technique was able to detect signals with lowest signal-to-noise-density ratio without error. In laboratory tests using GAUSS channel conditions, a system threshold of 13 dB-Hz could be demonstrated. In a real environment (North Cape) 15 dB-Hz was the lower limit for error-free reception. CCIR approved a recommendation in June, 1984 for a system operating through geostationary satellites at 1.6 GHz being a DRCS-type with very little modifications.

Cooperative communication has been considered a promising technique used in wireless network to introduce diversity. Based on distributed architecture, cooperative network enables single-antenna users to share their information as a virtual multiple antenna array, which makes it possible to enhance the user data rates and obtain the diversity gain. Depending on the ways users share and process the information, practical user cooperation is mainly grouped in three modes: Amplify and Forward (AF), Decode and Forward (DF) and Compress and Forward (CF). Generally, DF outperforms AF in most practical implementation and is alternately the best by theoretical analysis. In particular, it is shown in theory that DF outperforms CF when the inter-user link is in good condition and vice versa. Considered as a capacity achieving approach, superposition encoding technique produces approximate Gaussian output signals. It is the power and spectral efficiency of the encoding scheme that make it useful for several wireless communication systems such as MIMO system, relay channels, and so on. The main goal of the thesis is to study how much superposition coding can help when applied to DF user cooperation. In particular, a new DF scheme with soft information relaying technique is proposed to overcome the performance degradation happening when the quality of the inter-user channel is bad. Based on the scheme, soft log-likelihood ratio (LLR) of the relay decoder is quantized, encoded, modulated by superimposition and then forwarded to the destination. The destination receiver employs an iterative superposition decoder and modified maximum a posteriori probability (MAP) decoder to recover the transmitted signal from the source. The performance of the scheme is evaluated by simulation. We show that the new scheme can significantly outperform the conventional DF scheme even in a poor inter-user channel. Furthermore, a new cooperative transmission strategy based on superposition and multiuser detection techniques is also investigated in the thesis. The new scheme is an improved DF scheme in which the two users take turns in being the relay for each other and the signal forwarded by each user is the superimposed data of both users. Here, interleavers are introduced in the superposition process as an efficient user separation tool, providing a large improvement in error performance and also facilitating the decoding process at the destination. Different types of iterative receivers are examined. The performance of the scheme is evaluated by simulation. It is shown that the new transmission scheme outperforms the conventional one in terms of both spectral and power efficiency.

Thesis (Ph.D.)--University of Central Florida, 2008. / Advisers: Xin Li, Mubarak Shah. Includes bibliographical references (p. 127-130).

For this study, two analytical models were developed for predicting the depth-averaged velocity distribution (U) in trapezoidal, half-trapezoidal and rectangular channels with submerged and unsubmerged rigid cylinders. The first model uses linear superposition of momentum defects (MDS) and mass conservation, and is referred to as the MDS model. The second model uses linear superposition of velocity defects (VDS) and mass conservation, and is referred to as the VDS model. For implementing either the VDS or MDS model, a criterion is required for considering the wake created by an individual cylinder to be fully dissipated (i.e., a cutoff criterion). Also, implementing the MDS model requires numerical integration. Analyses were conducted to identify suitable cutoff criterion and an appropriate subinterval size for the numerical integration. Data from a physical model study conducted in a flume with a half-trapezoidal channel section was used to calibrate and validate the models. Data from a physical model study conducted in a rectangular channel section was also used to validate the model. Predicted values of U from the VDS and MDS models were within the range of ± 20 % of the trapezoidal channel section validation data. The models failed to accurately predict U for the rectangular channel section data. It is concluded that the models developed herein should be used only for half-trapezoidal channel sections. With respect to the trapezoidal channel section validation data, the MDS model yields a sum of squared errors that is 36% less than that yielded by the VDS model. Therefore, the MDS model is regarded as the best model overall for computing U in half-trapezoidal channel sections.

In this dissertation, we analyze Kolmogorov's superposition theorem for high dimensions. Our main goal is to explore and demonstrate the feasibility of an accurate implementation of Kolmogorov's theorem. First, based on Lorentz's ideas, we provide a thorough discussion on the proof and its numerical implementation of the theorem in dimension two. We present computational experiments which prove the feasibility of the theorem in applications of low dimensions (namely, dimensions two and three). Next, we present high dimensional extensions with complete and detailed proofs and provide the implementation that aims at applications with high dimensionality. The amalgamation of these ideas is evidenced by applications in image (two dimensional) and video (three dimensional) representations, the content based image retrieval, video retrieval, de-noising and in-painting, and Bayesian prior estimation of high dimensional data from the fields of computer vision and image processing.