Página 1 dos resultados de 471 itens digitais encontrados em 0.030 segundos

Improving an image annotation and retrieval agent using commonsense and personal knowledge; Improving an image retrieval agent using commonsense and personal knowledge

Waters, Kimberly G. (Kimberly Geneva), 1977-
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 54 p.; 3772931 bytes; 3772686 bytes; application/pdf; application/pdf
ENG
Relevância na Pesquisa
75.68%
by Kimberly G. Waters.; Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2002.; Includes bibliographical references (p. 45-46).

Query expansion techniques for question answering

Bilotti, Matthew W. (Matthew William), 1981-
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 109 p.; 5751699 bytes; 5764839 bytes; application/pdf; application/pdf
EN_US
Relevância na Pesquisa
65.68%
Query expansion is a technique used to boost performance of a document retrieval engine, such as those commonly found in question answering (QA) systems. Common methods of query expansion for Boolean keyword-based document retrieval engines include inserting query terms, such as alternate inflectional or derivational forms generated from existing query terms, or dropping query terms that are, for example, deemed to be too restrictive. In this thesis, I present a quantitative evaluation against a test; collection of my own design of five query expansion techniques, two term expansion methods and three term-dropping strategies. I present results that show that there exist best-performing query expansion algorithms that can be ex- perimentally optimized for specific tasks. My findings pose questions that suggest interesting avenues for further study of query expansion algorithms.; by Matthew W. Bilotti.; Thesis (M. Eng.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2004.; Includes bibliographical references (p. 105-109).

MacNews : an interactive news retrieval service for the Macintosh

Segal, David A
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 37 leaves; 1852619 bytes; 1854503 bytes; application/pdf; application/pdf
EN_US
Relevância na Pesquisa
75.6%
by David Andrew Segal.; Thesis (B.S.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 1989.; Includes bibliographical references (leaf 37).

Optoelectronic fiber webs for imaging applications

Arnold, Jerimy Reeves
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 73 p.; 2459428 bytes; 2462252 bytes; application/pdf; application/pdf
ENG
Relevância na Pesquisa
65.61%
We demonstrate the use of novel visible and infrared light-sensitive optoelectronic fiber in the development of large scale photodector arrays. Unlike conventional point photodetectors these one-dimensional linear photodectors are capable of sensing light along the entire length of the fiber and 360 radially. Multiple fibers can be arranged in an orthogonal grid to create a two-dimensional fiber web. The fiber web is capable of tracking a time-and space-varying beam, and output it onto a computer screen. Other imaging applications for the fiber web include image recovery for 2D images based on Computed Axial Tomography concepts. and lensless imaging. Lensless imaging is accomplished d using two fiber webs separated by a fixed distance, recovering the intensity distribution on each fiber web, and applying a phase retrieval algorithm to the two distributions. Furthermore, fiber webs consisting of six planar arrays forming a cube can be used to detect incident light in three dimensions.; by Jerimy Reeves Arnold.; Thesis (M. Eng.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2005.; Includes bibliographical references (p. 73).

Patent semantics : analysis, search and visualization of large text corpora

Lucas, Christopher G
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 48 leaves; 2507435 bytes; 2507882 bytes; application/pdf; application/pdf
ENG
Relevância na Pesquisa
65.68%
Patent Semantics is system for processing text documents by extracting features capturing their semantic content, and searching, clustering, and relating them by those same features. It is set apart from existing methodologies by combining a visualization scheme that integrates retrieval and clustering, providing a variety of ways to find and relate documents depending on their goals. In addition, the system provides an explanatory mechanism that makes the retrieval an understandable process rather than a black box. The domain in which the system currently works is biochemistry and molecular biology patents but it is not intrinsically constrained to any document set.; by Christopher G. Lucas.; Thesis (M. Eng. and S.B.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2004.; Includes bibliographical references (leaves 47-48).

Inverse problems in electromagnetics

Chen, Xudong, 1977-
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 165 p.; 8148284 bytes; 8155248 bytes; application/pdf; application/pdf
ENG
Relevância na Pesquisa
65.77%
Two inverse problems in electromagnetics are investigated in this thesis. The first is the retrieval of the effective constitutive parameters of metamaterials from the measurement of the reflection and the transmission coefficients. A robust method is proposed for the retrieval of metamaterials as isotropic media, and four improvements over the existing methods make the retrieval results more stable. Next, a new scheme is presented for the retrieval of a specific bianisotropic metamaterial that consists of split-ring resonators, which signifies that the cross polarization terms of the metamaterial are quantitatively investigated for the first time. Finally, an optimization approach is designed to achieve the retrieval of general bianisotropic media with 72 unknown parameters. The hybrid algorithm combining the differential evolution (DE) algorithm and the simplex method steadily converges to the exact solution. The second inverse problem deals with the detection and the classification of buried metallic objects using electromagnetic induction (EMI). Both the exciting and the induced magnetic fields are expanded as a linear combination of basic modes in the spheroidal coordinate system. Due to the orthogonality and the completeness of the spheroidal basic modes...

Synthetic Aperture Radar Interferometry with 3 satellites; InSAR with 3 satellites

Wong, Wallace D. (Wallace Dazheng)
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 128 p.; 6805700 bytes; 6811039 bytes; application/pdf; application/pdf
ENG
Relevância na Pesquisa
65.73%
Our study investigates interferometric SAR (InSAR) post-processing height retrieval techniques. We explore the possible improvements by adding a third satellite to the two already in orbit, and examine some potential uses of this setup. As such, we investigate three methods for height retrieval and compare their results with the original 2-satellite method. The first approach is data averaging; a simple method that extends from the results obtained using the 2-satellite method. The 3 sets of data obtained per sampling look are grouped into pairs, and the 2 statistical best pairs are selected to be averaged, producing a better estimate of the digital elevation map (DEM) height. The second approach is the unambiguous range magnification (URM) method, which seeks to ease the reliance on phase unwrapping steps often necessary in retrieving height. It does so by expanding the wrapped phase range without performing any phase unwrapping, through the use of different wrapping speeds of the 3 sets of satellite pairings. The third method is the maximum likelihood estimation technique, an asymptotically efficient method which employs the same phase expansion property as the URM to predict the closest phase estimate which best fits most (if not all) of the data sets provided.; (cont.) Results show that for a handful of flyover looks...

Measurement and device design of left-handed metamaterials

Thomas Zachary M. (Zachary Michael)
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 128 p.; 5816207 bytes; 5821581 bytes; application/pdf; application/pdf
ENG
Relevância na Pesquisa
65.6%
The properties of a variety of left-handed metamaterial (LHM) structures are analyzed and measured to verify consistent behavior between theory an measurements. The structures are simulated using a commercial software program and a retrieval algorithm is used to determine the effective constitutive parameters. The constitutive parameters are used to predict the behavior of the metamaterial under various configurations. Measurements are conducted to verify the presence of a negative index of refraction. Transmission through an LHM slab from several incidences is shown to be consistent with theory. A four-port device utilizing the dispersive nature of an LHM prism is designed and measured. The measurements show that the refraction angle of an incident signal is frequency dependent. Two ports are constructed to receive the positively refracted and negatively refracted power. In the frequency band where the incident signal cannot propagate in the LHM prism, the power is reflected from the interface towards a third measurement port. The three ports are shown to achieve unique mutually exclusive bandwidths. A general study is conducted on the design of such a device. Finally, the use of a left-handed metamaterial as a substrate for a microstrip line is investigated.; (cont.) An LHM substrate consisting of split-ring resonators is shown to enhance the performance of a stop band filter. The measurement results are in good agreement with simulation where the substrate is modelled by its effective medium parameters.; by Zachary M. Thomas.; Thesis (S.M.)--Massachusetts Institute of Technology...

OverCite : a cooperative digital research library; Over Cite : a cooperative digital research library

Stribling, Jeremy
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 50 leaves; 2496775 bytes; 2497462 bytes; application/pdf; application/pdf
ENG
Relevância na Pesquisa
65.62%
CiteSeer is a well-known online resource for the computer science research community, allowing users to search and browse a large archive of research papers. Unfortunately, its current centralized incarnation is costly to run. Although members of the community would presumably be willing to donate hardware and bandwidth at their own sites to assist CiteSeer, the current architecture does not facilitate such distribution of resources. OverCite is a design for a new architecture for a distributed and cooperative research library based on a distributed hash table (DHT). The new architecture harnesses donated resources at many sites to provide document search and retrieval service to researchers worldwide. A preliminary evaluation of an initial OverCite prototype shows that it can service more queries per second than a centralized system, and that it increases total storage capacity by a factor of n/4 in a system of n nodes. OverCite can exploit these additional resources by supporting new features such as document alerts, and by scaling to larger data sets.; by Jeremy Stribling.; Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2005.; Includes bibliographical references (leaves 47-50).

Matching sets of features for efficient retrieval and recognition

Grauman, Kristen Lorraine, 1979-
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 153 p.
ENG
Relevância na Pesquisa
75.74%
In numerous domains it is useful to represent a single example by the collection of local features or parts that comprise it. In computer vision in particular, local image features are a powerful way to describe images of objects and scenes. Their stability under variable image conditions is critical for success in a wide range of recognition and retrieval applications. However, many conventional similarity measures and machine learning algorithms assume vector inputs. Comparing and learning from images represented by sets of local features is therefore challenging, since each set may vary in cardinality and its elements lack a meaningful ordering. In this thesis I present computationally efficient techniques to handle comparisons, learning, and indexing with examples represented by sets of features. The primary goal of this research is to design and demonstrate algorithms that can effectively accommodate this useful representation in a way that scales with both the representation size as well as the number of images available for indexing or learning. I introduce the pyramid match algorithm, which efficiently forms an implicit partial matching between two sets of feature vectors.; (cont.) The matching has a linear time complexity...

Optical and optoelectronic fiber devices

Shapira, Ofer, Ph. D. Massachusetts Institute of Technology
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 119 p.
ENG
Relevância na Pesquisa
65.68%
The ability to integrate materials with disparate electrical, thermal, and optical properties into a single fiber structure enabled the realization of fiber devices with diverse and complex functionalities. Amongst those, demonstrated first in our work, are the surface-emitting fiber laser, the hollow-core fiber amplifier, the thermally self-monitored high-power transmission fiber device, and the photo-detecting fiber-web based imaging system. This work presents the design, analysis, and characterization of those devices. It opens with a study of the transmission properties of the multimode hollow-core, photonic bandgap fiber constructed of a periodic multilayer cladding. A defect is then introduced into one of the cladding layers and the interaction between core and defect modes is investigated. The second chapter addresses the experimental problem encountered in many multimode waveguide applications: how to extract, and to some extent to control, the modal content of the field at the output of a waveguide. We developed a non-interferometric approach to achieve mode decomposition based on a modified phase retrieval algorithm that can yield the complete vectorial eigenmode content of any general waveguiding structure and demonstrated its validity experimentally. In the third chapter an active material is introduced into the hollow-core to form a surface-emitting fiber laser. A unique azimuthally anisotropic optical wave front results from the interplay between the cylindrical resonator...

Photo annotation and retrieval through speech

Sherry, Brennan P
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 36 p.
ENG
Relevância na Pesquisa
75.77%
In this thesis I describe the development of a speech-based annotation and retrieval system for digital photographs. The system uses a client/server architecture which allows photographs to be captured and annotated on various clients, such as mobile camera phones or the web, and then processed, indexed and stored on networked servers. For speech-based retrieval we have developed a mixed grammar recognition approach which allows the speech recognition system to construct a single finite-state network combining context-free grammars for recognizing and parsing query carrier phrases and metadata phrases, with an unconstrained statistical n-gram model for recognizing free-form search terms. Experiments demonstrating successful retrieval of photographs using purely speech-based annotation and retrieval are presented.; by Brennan P. Sherry.; Thesis (M. Eng.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2007.; Includes bibliographical references (p. 36).

Efficient object recognition and image retrieval for large-scale applications

Lee, John Jaesung
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 93 p.
ENG
Relevância na Pesquisa
75.7%
Algorithms for recognition and retrieval tasks generally call for both speed and accuracy. When scaling up to very large applications, however, we encounter additional significant requirements: adaptability and scalability. In many real-world systems, large numbers of images are constantly added to the database, requiring the algorithm to quickly tune itself to recent trends so it can serve queries more effectively. Moreover, the systems need to be able to meet the demands of simultaneous queries from many users. In this thesis, I describe two new algorithms intended to meet these requirements and give an extensive experimental evaluation for both. The first algorithm constructs an adaptive vocabulary forest, which is an efficient image-database model that grows and shrinks as needed while adapting its structure to tune itself to recent trends. The second algorithm is a method for efficiently performing classification tasks by comparing query images to only a fixed number of training examples, regardless of the size of the image database. These two methods can be combined to create a fast, adaptable, and scalable vision system suitable for large-scale applications. I also introduce LIBPMK, a fast implementation of common computer vision processing pipelines such as that of the pyramid match kernel. This implementation was used to build several successful interactive applications as well as batch experiments for research settings. This implementation...

Mode conversation losses in overmolded millimeter wave transmission lines

Tax, David S. (David Samuel)
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 109 p.
ENG
Relevância na Pesquisa
65.6%
Millimeter wave transmission lines are integral components for many important applications like nuclear fusion and NMR spectroscopy. In low loss corrugated transmission lines propagating the HE,1 mode with a high waveguide radius to wavelength ratio (a/X), the transmission line loss is predominantly a result of mode conversion in components such as miter bends. The theory for determining losses in miter bends though is only approximate, and is based instead on the problem of the loss across a diameter-length gap between two waveguide sections. Through simulation, we verified that the existing analytic theory of this gap loss is correct; however, our simulations could not verify the assumption that the miter bend loss is half the loss in the gap. We also considered the problem of higher order modes (HOMs) mixed with an HE11 input entering the miter bend. Using a numerical technique, we found that the loss through the miter bend is dependent on both the amplitude of the HOM content as well as its phase relative to the phase of the HE11 mode. While the overall loss averaged across all phases remains the same with increasing HOM content, the power that fails to traverse the gap tends to increase, and it is this power that appears as very high order modes that will cause heating around the miter bend. For the ITER transmission line...

The design and implementation of a parallel document retrieval engine

Hawking, David
Fonte: Universidade Nacional da Austrália Publicador: Universidade Nacional da Austrália
Tipo: Working/Technical Paper Formato: 263051 bytes; 356 bytes; application/pdf; application/octet-stream
EN_AU
Relevância na Pesquisa
55.66%
Document retrieval as traditionally formulated is an inherently parallel task because the document collection can be divided into N sub-collections each of which may be searched independently. Document retrieval software can potentially exploit the power and capacity of a large-scale parallel machine to improve speed, to extend the size of the largest collection which can be processed, to respond quickly to changes in the document collection and/or to increase the power and expressivity of the retrieval query language. This paper includes discussion of the issues involved in the design of a practical parallel document retrieval engine for a distributed-memory multicomputer and a description of the implementation of PADRE, a retrieval engine for the Fujitsu AP1000. Performance results are presented and scope of applicability of the techniques is discussed.; no

Multi-baseline interferometric synthetic aperture radar applications and error analysis; Multi-baseline interferometric SAR applications and error analysis

Chua, Song Liang
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 130 p.
ENG
Relevância na Pesquisa
65.81%
In this thesis, we deal primarily with the multi-baseline SAR configuration utilizing three satellites. Two applications of InSAR, multi-baseline height retrieval and multi-baseline compensation of CCD's slope biasing effects, are first examined in details. An optimal baseline-weighted height averaging technique is introduced. Phase averaging, a novel height retrieval technique, combines the multi-baseline phase data into one, such that only one set of heights is retrieved from the three-satellite configuration. This approach outperforms single baseline height retrieval and allows application of the conventional two-satellite height retrieval process on the multi-baseline data, without need for excessive modifications. Slope biasing effects, inherent in multilook coherence estimator, make it difficult to identify if low or medium coherence values are results of an actual scene change or an undulating terrain. This ambiguity can be best resolved by accounting for the topographic phase variations via prior knowledge of the original height profile, whose precise retrieval requires a multi-baseline satellite configuration. The three-satellite setup is then related to a realistic cartwheel configuration, where the resulting errors in the height retrieval and CCD process...

Lower bounds for embedding the Earth Mover Distance metric into normed spaces

Samuel, Javed K. K
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 73 p.; 1891353 bytes; 1894338 bytes; application/pdf; application/pdf
ENG
Relevância na Pesquisa
65.6%
This thesis presents a lower bounds for embedding the Earth Mover Distance (EMID) metric into normed spaces. The EMID is a metric over two distributions where one is a mass of earth spread out in space and the other is a collection of holes in that same space. The EMD between these two distributions is defined as the least amount of work needed to fill the holes with earth. The EMD metric is used in a number of applications, for example in similarity searching and for image retrieval. We present a simple construction of point sets in the ENID metric space over two dimensions that cannot be embedded from the ED metric exactly into normed spaces, namely l1 and the square of l2. An embedding is a mapping f : X --> V with X a set of points in a metric space and ' Va set of points in some normed vector space. When the Manhattan distance is used as the underlying metric for the EMD, it can be shown that this example is isometric to K2,4 which has distortion equal to 1.25 when it is embedded into I and( 1.1180 when embedded into the square of 12. Other constructions of points sets in the EMID metric space over three and higher dimensisions are also discussed..; by Javed K.K. Samuel.; Thesis (M. Eng.)--Massachusetts Institute of Technology...

Geophysical parameter estimation with a passive microwave spectrometer at 54 / 118 / 183 / 425 GHz

Leslie, Robert Vincent, 1972-
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 271 p.; 13073198 bytes; 13109168 bytes; application/pdf; application/pdf
EN_US
Relevância na Pesquisa
65.68%
(cont.) model of a convective cell is presented that provides a physical basis for this relationship.; The National Polar-orbiting Operational Environmental Satellite System (NPOESS) Aircraft Sounder Testbed-Microwave, or NAST-M, includes passive microwave spectrometers operating near the oxygen lines at 50-57 GHz, 118.75 GHz, and 424.76 GHz, and a spectrometer centered on the water vapor absorption line at 183.31 GHz. All four of the spectrometers' antenna horns are colocated, have 3-dB (full-width at half-max) beamwidths of 7.5⁰, and are directed at a single mirror that scans cross-track beneath the aircraft with a swath up to 100-km wide. The 183.31- and 424.76-GHz systems were developed as part of this thesis. The calibration techniques for two high-altitude airborne platforms are described and validated for two recent deployments. During these two deployments, various precipitation phenomena were imaged by NAST-M's radiometric and video instruments. Retrieval methods were developed and tested for single-pixel rain rate, precipitation cell-top altitude, and cloud-top altitude retrievals of convective cells. The basis of the single-pixel retrievals is a simplified convective-cell hydrometeor-profile model used with a radiative transfer solution that included absorption by atmospheric gases and by hydrometeor absorption and scattering. Two retrieval techniques were used to relate the simulated brightness temperatures to the actual brightness temperatures from the deployments. Case studies are presented from each deployment. In addition...

Retrieval of cloud-cleared atmospheric temperature profiles from hyperspectral infrared and microwave observations

Blackwell, William Joseph, 1971-
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 307 p.; 10143420 bytes; 10143219 bytes; application/pdf; application/pdf
ENG
Relevância na Pesquisa
75.75%
This thesis addresses the problem of retrieving the temperature profile of the Earth's atmosphere from overhead infrared and microwave observations of spectral radiance in cloudy conditions. The contributions of the thesis are twofold: improvements in 1) microwave instrumentation and 2) hyperspectral signal processing and estimation algorithms. The NPOESS Aircraft Sounder Testbed-Microwave (NAST-M) passive spectrometer was designed, fabricated and deployed. NAST-M provides accurate brightness temperature measurements in 16 channels near the oxygen absorption lines at 50-57 GHz and 118.75 GHz, permitting the first reliably accurate retrieval images of temperature profiles and precipitation structure in cloudy areas. The correlation structure of the NPOESS Aircraft Sounder Testbed-Infrared (NAST-I) instrument noise was analyzed in the spectral and spatial domains using the Iterated Order-Noise (ION) algorithm [1] for two representative flights. Results indicate that vibration-induced noise was the dominant component, but that it could be significantly reduced by filtering in the spatial domain. Novel multi-pixel cloud clearing and temperature profile retrieval algorithms were developed for simulated Atmospheric Infrared Sounder (AIRS) and Advanced Microwave Sounding Unit (AMSU) radiances using neural networks. RMS temperature profile retrieval errors of -0.5 K were obtained for all levels of the atmosphere from 0-15 km in clear air at a horizontal resolution of 2000 km2 and a vertical resolution of 1 km. RMS radiance errors under cloudy conditions for altitudes from 0 to 10 kilometers ranged from 1.25 K to 0.1 K for radiance retrievals near 15 microns...

Evolutionary Computational Method of Facial Expression Analysis for Content-based Video Retrieval using 2-Dimensional Cellular Automata

Geetha, P.; Narayanan, Vasumathi
Fonte: Universidade Cornell Publicador: Universidade Cornell
Tipo: Artigo de Revista Científica
Publicado em 10/09/2010
Relevância na Pesquisa
55.7%
In this paper, Deterministic Cellular Automata (DCA) based video shot classification and retrieval is proposed. The deterministic 2D Cellular automata model captures the human facial expressions, both spontaneous and posed. The determinism stems from the fact that the facial muscle actions are standardized by the encodings of Facial Action Coding System (FACS) and Action Units (AUs). Based on these encodings, we generate the set of evolutionary update rules of the DCA for each facial expression. We consider a Person-Independent Facial Expression Space (PIFES) to analyze the facial expressions based on Partitioned 2D-Cellular Automata which capture the dynamics of facial expressions and classify the shots based on it. Target video shot is retrieved by comparing the similar expression is obtained for the query frame's face with respect to the key faces expressions in the database video. Consecutive key face expressions in the database that are highly similar to the query frame's face, then the key faces are used to generate the set of retrieved video shots from the database. A concrete example of its application which realizes an affective interaction between the computer and the user is proposed. In the affective interaction, the computer can recognize the facial expression of any given video shot. This interaction endows the computer with certain ability to adapt to the user's feedback.; Comment: Submitted to Journal of Computer Science and Engineering...