（1）ICCV1999 Object Recognition from Local Scale-Invariant Features.pdf提出（2）IJCV2004 Distinctive image features from scale invariant keypoints.pdf總結（3）CVPR2004 PCA-SIFT：A More Distinctive REPRESENTATION for Local Image Descriptors.pdf加PCA降維
In this project we analyze and design the minimum mean-square error (MMSE) multiuser receiver for uniformly quantized synchronous code division multiple access (CDMA) signals in additive white Gaussian noise (AWGN) channels.This project is mainly based on the REPRESENTATION of uniform quantizer by gain plus additive noise model. Based on this model, we derive the weight vector and the output signal-to-interference ratio (SIR) of the MMSE receiver. The effects of quantization on the MMSE receiver performance is characterized in a single parameter named 鈥漞quivalent noise variance鈥? The optimal quantizer stepsize which maximizes the MMSE receiver output SNR is also determined.
H.264/AVC is the current video standardization project of the ITU-T Video Coding
Experts Group (VCEG) and the ISO/IEC Moving Picture Experts Group (MPEG). The
main goals of this standardization effort are to develop a simple and straightforward
video coding design, with enhanced compression performance, and to provide a
鈥渘etwork-friendly鈥?video REPRESENTATION which addresses 鈥渃onversational鈥?(video
telephony) and 鈥渘on-conversational鈥?(storage, broadcast or streaming) applications.
In the last decade the processing of polygonal meshes has
emerged as an active and very productive research area. This
can basically be attributed to two developments:
Modern geometry acquisition devices, like laser scanners
and MRT, easily produce raw polygonal meshes of
ever growing complexity
Downstream applications like analysis tools (medical
imaging), computer aided manufacturing, or numerical
simulations all require high quality polygonal meshes
as input.
The need to bridge the gap between raw triangle soup data
and high-quality polygon meshes has driven the research on
ecient data structures and algorithms that directly operate
on polygonal meshes rather than on a (most often not
feasible) intermediate CAD REPRESENTATION.
SFLOP simulates a floating point operation
x op y where op = +, -, *, /
In chopping or rounding arithmetic using an
m digit mantissa, base 10, and an unrestricted
exponent range. (sflop: Simulate FLOating Point
operation.) For more details on the how the floating
point REPRESENTATION of a number is computed see
the function sfl.
Text mining tries to solve the crisis of information overload by combining techniques from data mining, machine learning, natural language processing, information retrieval, and knowledge management. In addition to providing an in-depth examination of core text mining and link detection algorithms and operations, this book examines advanced pre-processing techniques, knowledge REPRESENTATION considerations, and visualization approaches. Finally, it explores current real-world, mission-critical applications of text mining and link detection in such varied fields as M&A business intelligence, genomics research and counter-terrorism activities.
Abstract—We describe a technique for image encoding in which local operators of many scales but identical shape serve as the basis functions. The REPRESENTATION differs from established techniques in that the code elements are localized in spatial frequency as well as in space.
The algorithm ID3 (Quinlan) uses the method top-down induction of decision trees. Given a set of classified examples a decision tree is induced, biased by the information gain measure, which heuristically leads to small trees. The examples are given in attribute-value REPRESENTATION. The set of possible classes is finite. Only tests, that split the set of instances of the underlying example languages depending on the value of a single attribute are supported.
The object detector described below has been initially proposed by
P.F. Felzenszwalb in [Felzenszwalb2010]. It is based on a
Dalal-Triggs detector that uses a single filter on histogram of
oriented gradients (HOG) features to represent an object category.
This detector uses a sliding window approach, where a filter is
applied at all positions and scales of an image. The first
innovation is enriching the Dalal-Triggs model using a
star-structured part-based model defined by a “root” filter
(analogous to the Dalal-Triggs filter) plus a set of parts filters
and associated deformation models. The score of one of star models
at a particular position and scale within an image is the score of
the root filter at the given location plus the sum over parts of the
maximum, over placements of that part, of the part filter score on
its location minus a deformation cost easuring the deviation of the
part from its ideal location relative to the root. Both root and
part filter scores are defined by the dot product between a filter
(a set of weights) and a subwindow of a feature pyramid computed
from the input image. Another improvement is a REPRESENTATION of the
class of models by a mixture of star models. The score of a mixture
model at a particular position and scale is the maximum over
components, of the score of that component model at the given
location.
The TJA1040 is an advanced high speed CAN transceiver for use in
automotive and general industrial applications. It supports the differential
bus signal REPRESENTATION described in the international standard for
in-vehicle high speed CAN applications (ISO11898). CAN (Controller Area
Network) is the standard protocol for serial in-vehicle bus communication,
particularly for Engine Management and Body Multiplexing.
The TJA1040 provides a Standby mode, as known from its functional
predecessors PCA82C250 and PCA82C251, but with significantly
reduced power consumption. Besides the excellent low-power behavior
the TJA1040 offers several valuable system improvements. Highlights are
the absolute passive bus behavior if the device is unpowered as well as
the excellent EMC performance.
