TY - CONF
T1 - A large-scale benchmark dataset for event recognition in surveillance video
T2 - Computer Vision and Pattern Recognition (CVPR), 2011 IEEE Conference on
Y1 - 2011
A1 - Oh,Sangmin
A1 - Hoogs, A.
A1 - Perera,A.
A1 - Cuntoor, N.
A1 - Chen,Chia-Chih
A1 - Lee,Jong Taek
A1 - Mukherjee,S.
A1 - Aggarwal, JK
A1 - Lee,Hyungtae
A1 - Davis, Larry S.
A1 - Swears,E.
A1 - Wang,Xioyang
A1 - Ji,Qiang
A1 - Reddy,K.
A1 - Shah,M.
A1 - Vondrick,C.
A1 - Pirsiavash,H.
A1 - Ramanan,D.
A1 - Yuen,J.
A1 - Torralba,A.
A1 - Song,Bi
A1 - Fong,A.
A1 - Roy-Chowdhury, A.
A1 - Desai,M.
KW - algorithm;evaluation
KW - CVER
KW - databases;
KW - databases;video
KW - dataset;moving
KW - event
KW - metrics;large-scale
KW - object
KW - recognition
KW - recognition;diverse
KW - recognition;video
KW - scenes;surveillance
KW - surveillance;visual
KW - tasks;computer
KW - tracks;outdoor
KW - video
KW - video;computer
KW - vision;continuous
KW - vision;image
KW - visual
AB - We introduce a new large-scale video dataset designed to assess the performance of diverse visual event recognition algorithms with a focus on continuous visual event recognition (CVER) in outdoor areas with wide coverage. Previous datasets for action recognition are unrealistic for real-world surveillance because they consist of short clips showing one action by one individual [15, 8]. Datasets have been developed for movies [11] and sports [12], but, these actions and scene conditions do not apply effectively to surveillance videos. Our dataset consists of many outdoor scenes with actions occurring naturally by non-actors in continuously captured videos of the real world. The dataset includes large numbers of instances for 23 event types distributed throughout 29 hours of video. This data is accompanied by detailed annotations which include both moving object tracks and event examples, which will provide solid basis for large-scale evaluation. Additionally, we propose different types of evaluation modes for visual recognition tasks and evaluation metrics along with our preliminary experimental results. We believe that this dataset will stimulate diverse aspects of computer vision research and help us to advance the CVER tasks in the years ahead.
JA - Computer Vision and Pattern Recognition (CVPR), 2011 IEEE Conference on
M3 - 10.1109/CVPR.2011.5995586
ER -
TY - CONF
T1 - Modeling temporal correlations in content fingerprints
T2 - Acoustics, Speech and Signal Processing (ICASSP), 2011 IEEE International Conference on
Y1 - 2011
A1 - Varna,A.L.
A1 - M. Wu
KW - chain
KW - correlations;video
KW - database;Markov
KW - databases;
KW - databases;video
KW - detection;fingerprint
KW - detector;certain
KW - fingerprints;hybrid
KW - identification;temporal
KW - Markov
KW - model;adaptive
KW - model;temporal
KW - processes;adaptive
KW - regime;content
KW - signal
AB - Previous analysis of content fingerprints has mainly focused on the case of independent and identically distributed finger prints. Practical fingerprints, however, exhibit correlations between components computed from successive frames. In this paper, a Markov chain based model is used to capture the temporal correlations, and the suitability of this model is evaluated through experiments on a video database. The results indicate that the Markov chain model is a good fit only in a certain regime. A hybrid model is then developed to account for this behavior and a corresponding adaptive detector is derived. The adaptive detector achieves better identification accuracy at a small computational expense.
JA - Acoustics, Speech and Signal Processing (ICASSP), 2011 IEEE International Conference on
M3 - 10.1109/ICASSP.2011.5946868
ER -
TY - CONF
T1 - Evaluation of state-of-the-art algorithms for remote face recognition
T2 - Image Processing (ICIP), 2010 17th IEEE International Conference on
Y1 - 2010
A1 - Ni,Jie
A1 - Chellapa, Rama
KW - alarm;feature
KW - algorithm;still
KW - classification;image
KW - classification;visual
KW - database;remote
KW - databases;
KW - extraction;hidden
KW - extraction;image
KW - Face
KW - feature
KW - image-based
KW - image;false
KW - quality;occlusion;remote
KW - recognition;face
KW - recognition;feature
KW - recognition;state-of-the-art
KW - removal;image
AB - In this paper, we describe a remote face database which has been acquired in an unconstrained outdoor environment. The face images in this database suffer from variations due to blur, poor illumination, pose, and occlusion. It is well known that many state-of-the-art still image-based face recognition algorithms work well, when constrained (frontal, well illuminated, high-resolution, sharp, and complete) face images are presented. In this paper, we evaluate the effectiveness of a subset of existing still image-based face recognition algorithms for the remote face data set. We demonstrate that in addition to applying a good classification algorithm, consistent detection of faces with fewer false alarms and finding features that are robust to variations mentioned above are very important for remote face recognition. Also setting up a comprehensive metric to evaluate the quality of face images is necessary in order to reject images that are of low quality.
JA - Image Processing (ICIP), 2010 17th IEEE International Conference on
M3 - 10.1109/ICIP.2010.5652608
ER -
TY - JOUR
T1 - Face Verification Across Age Progression Using Discriminative Methods
JF - Information Forensics and Security, IEEE Transactions on
Y1 - 2010
A1 - Ling,Haibin
A1 - Soatto,S.
A1 - Ramanathan,N.
A1 - Jacobs, David W.
KW - algorithms;support
KW - databases;
KW - dataset;age
KW - FGnet
KW - hair;gradient
KW - information;image
KW - information;recognition
KW - machine;face
KW - machines;visual
KW - methods;eyewear;face
KW - methods;support
KW - orientation
KW - orientation;gradient
KW - progression;commercial
KW - pyramid;hierarchical
KW - quality;magnitude
KW - recognition;gradient
KW - systems;discriminative
KW - vector
KW - verification;facial
AB - Face verification in the presence of age progression is an important problem that has not been widely addressed. In this paper, we study the problem by designing and evaluating discriminative approaches. These directly tackle verification tasks without explicit age modeling, which is a hard problem by itself. First, we find that the gradient orientation, after discarding magnitude information, provides a simple but effective representation for this problem. This representation is further improved when hierarchical information is used, which results in the use of the gradient orientation pyramid (GOP). When combined with a support vector machine GOP demonstrates excellent performance in all our experiments, in comparison with seven different approaches including two commercial systems. Our experiments are conducted on the FGnet dataset and two large passport datasets, one of them being the largest ever reported for recognition tasks. Second, taking advantage of these datasets, we empirically study how age gaps and related issues (including image quality, spectacles, and facial hair) affect recognition algorithms. We found surprisingly that the added difficulty of verification produced by age gaps becomes saturated after the gap is larger than four years, for gaps of up to ten years. In addition, we find that image quality and eyewear present more of a challenge than facial hair.
VL - 5
SN - 1556-6013
CP - 1
M3 - 10.1109/TIFS.2009.2038751
ER -
TY - CONF
T1 - Geotagging with local lexicons to build indexes for textually-specified spatial data
T2 - Data Engineering (ICDE), 2010 IEEE 26th International Conference on
Y1 - 2010
A1 - Lieberman,M.D.
A1 - Samet, Hanan
A1 - Sankaranarayanan,J.
KW - data;Internet;data
KW - databases;
KW - databases;spatial
KW - document
KW - indexes;spatial
KW - information
KW - Internet;document-independent
KW - knowledge;feature-based
KW - lexicon
KW - lexicons;location-based
KW - locations;geotagging;inference
KW - method;internal
KW - methods;geographic
KW - mining;geographic
KW - model;external
KW - model;textually-specified
KW - queries;generic
KW - queries;spatial
KW - spatial
KW - structure;local
KW - systems;visual
AB - The successful execution of location-based and feature-based queries on spatial databases requires the construction of spatial indexes on the spatial attributes. This is not simple when the data is unstructured as is the case when the data is a collection of documents such as news articles, which is the domain of discourse, where the spatial attribute consists of text that can be (but is not required to be) interpreted as the names of locations. In other words, spatial data is specified using text (known as a toponym) instead of geometry, which means that there is some ambiguity involved. The process of identifying and disambiguating references to geographic locations is known as geotagging and involves using a combination of internal document structure and external knowledge, including a document-independent model of the audience's vocabulary of geographic locations, termed its spatial lexicon. In contrast to previous work, a new spatial lexicon model is presented that distinguishes between a global lexicon of locations known to all audiences, and an audience-specific local lexicon. Generic methods for inferring audiences' local lexicons are described. Evaluations of this inference method and the overall geotagging procedure indicate that establishing local lexicons cannot be overlooked, especially given the increasing prevalence of highly local data sources on the Internet, and will enable the construction of more accurate spatial indexes.
JA - Data Engineering (ICDE), 2010 IEEE 26th International Conference on
M3 - 10.1109/ICDE.2010.5447903
ER -
TY - CONF
T1 - Action recognition based on human movement characteristics
T2 - Motion and Video Computing, 2009. WMVC '09. Workshop on
Y1 - 2009
A1 - Dondera,R.
A1 - David Doermann
A1 - Davis, Larry S.
KW - action
KW - ballistic
KW - characteristics;motion
KW - correlated
KW - cost;human
KW - data;probability
KW - databases;
KW - density
KW - descriptor;motion
KW - dynamics;computational
KW - function;robustness;shape
KW - information;short
KW - linear
KW - Movement
KW - movements;computer
KW - recognition;human
KW - recognition;stability;visual
KW - vector
KW - vision;pattern
AB - We present a motion descriptor for human action recognition where appearance and shape information are unreliable. Unlike other motion-based approaches, we leverage image characteristics specific to human movement to achieve better robustness and lower computational cost. Drawing on recent work on motion recognition with ballistic dynamics, an action is modeled as a series of short correlated linear movements and represented with a probability density function over motion vector data. We are targeting common human actions composed of ballistic movements, and our descriptor can handle both short actions (e.g. reaching with the hand) and long actions with events at relatively stable time offsets (e.g. walking). The proposed descriptor is used for both classification and detection of action instances, in a nearest-neighbor framework. We evaluate the descriptor on the KTH action database and obtain a recognition rate of 90% in a relevant test setting, comparable to the state-of-the-art approaches that use other cues in addition to motion. We also acquired a database of actions with slight occlusion and a human actor manipulating objects of various shapes and appearances. This database makes the use of appearance and shape information problematic, but we obtain a recognition rate of 95%. Our work demonstrates that human movement has distinctive patterns, and that these patterns can be used effectively for action recognition.
JA - Motion and Video Computing, 2009. WMVC '09. Workshop on
M3 - 10.1109/WMVC.2009.5399233
ER -
TY - CONF
T1 - Aggregate Query Answering under Uncertain Schema Mappings
T2 - Data Engineering, 2009. ICDE '09. IEEE 25th International Conference on
Y1 - 2009
A1 - Gal,A.
A1 - Martinez,M. V
A1 - Simari,G. I
A1 - V.S. Subrahmanian
KW - aggregate
KW - algorithm;probabilistic
KW - answering;by-table
KW - complexity;data
KW - complexity;distributed
KW - database;polynomial
KW - databases;
KW - databases;probability;query
KW - integration;distribution
KW - mapping;computational
KW - mapping;query
KW - processing;range
KW - processing;statistical
KW - query
KW - schema
KW - semantics;by-tuple
KW - semantics;computational
KW - semantics;expected
KW - semantics;multiple
KW - semantics;uncertain
KW - TIME
KW - value
AB - Recent interest in managing uncertainty in data integration has led to the introduction of probabilistic schema mappings and the use of probabilistic methods to answer queries across multiple databases using two semantics: by-table and by-tuple. In this paper, we develop three possible semantics for aggregate queries: the range, distribution, and expected value semantics, and show that these three semantics combine with the by-table and by-tuple semantics in six ways. We present algorithms to process COUNT, AVG, SUM, MIN, and MAX queries under all six semantics and develop results on the complexity of processing such queries under all six semantics. We show that computing COUNT is in PTIME for all six semantics and computing SUM is in PTIME for all but the by-tuple/distribution semantics. Finally, we show that AVG, MIN, and MAX are PTIME computable for all by-table semantics and for the by-tuple/range semantics.We developed a prototype implementation and experimented with both real-world traces and simulated data. We show that, as expected, naive processing of aggregates does not scale beyond small databases with a small number of mappings. The results also show that the polynomial time algorithms are scalable up to several million tuples as well as with a large number of mappings.
JA - Data Engineering, 2009. ICDE '09. IEEE 25th International Conference on
M3 - 10.1109/ICDE.2009.55
ER -
TY - JOUR
T1 - Moving Object Verification in Airborne Video Sequences
JF - Circuits and Systems for Video Technology, IEEE Transactions on
Y1 - 2009
A1 - Yue,Zhanfeng
A1 - Guarino, D.
A1 - Chellapa, Rama
KW - airborne
KW - database;homography-based
KW - databases;
KW - matcher;color
KW - matcher;image
KW - matching;image
KW - method;infrared
KW - object
KW - sequences;color
KW - sequences;distance
KW - sequences;moving
KW - sequences;video
KW - synthesis
KW - system;exemplar
KW - transforms;end-to-end
KW - verification;spatial-feature
KW - video
KW - view
AB - This paper presents an end-to-end system for moving object verification in airborne video sequences. Using a sample selection module, the system first selects frames from a short sequence and stores them in an exemplar database. To handle appearance change due to potentially large aspect angle variations, a homography-based view synthesis method is then used to generate a novel view of each image in the exemplar database at the same pose as the testing object in each frame of a testing video segment. A rotationally invariant color matcher and a spatial-feature matcher based on distance transforms are combined using a weighted average rule to compare the novel view and the testing object. After looping over all testing frames, the set of match scores is passed to a temporal analysis module to examine the behavior of the testing object, and calculate a final likelihood. Very good verification performance is achieved over thousands of trials for both color and infrared video sequences using the proposed system.
VL - 19
SN - 1051-8215
CP - 1
M3 - 10.1109/TCSVT.2008.2009243
ER -
TY - CONF
T1 - Secure image retrieval through feature protection
T2 - Acoustics, Speech and Signal Processing, 2009. ICASSP 2009. IEEE International Conference on
Y1 - 2009
A1 - Lu,Wenjun
A1 - Varna,A.L.
A1 - Swaminathan,A.
A1 - M. Wu
KW - bit-plane
KW - confidentiality;data
KW - database;image
KW - databases;
KW - encoding;signal
KW - extraction;image
KW - feature
KW - processes;visual
KW - processing;content-based
KW - projection;randomized
KW - protection;image
KW - randomization;cryptographic
KW - retrieval;cryptography;feature
KW - retrieval;random
KW - storage;encrypted
KW - technique;data
KW - unary
AB - This paper addresses the problem of image retrieval from an encrypted database, where data confidentiality is preserved both in the storage and retrieval process. The paper focuses on image feature protection techniques which enable similarity comparison among protected features. By utilizing both signal processing and cryptographic techniques, three schemes are investigated and compared, including bit-plane randomization, random projection, and randomized unary encoding. Experimental results show that secure image retrieval can achieve comparable retrieval performance to conventional image retrieval techniques without revealing information about image content. This work enriches the area of secure information retrieval and can find applications in secure online services for images and videos.
JA - Acoustics, Speech and Signal Processing, 2009. ICASSP 2009. IEEE International Conference on
M3 - 10.1109/ICASSP.2009.4959888
ER -
TY - JOUR
T1 - SPOT Databases: Efficient Consistency Checking and Optimistic Selection in Probabilistic Spatial Databases
JF - Knowledge and Data Engineering, IEEE Transactions on
Y1 - 2009
A1 - Parker,A.
A1 - Infantes,G.
A1 - Grant,J.
A1 - V.S. Subrahmanian
KW - Cartesian
KW - checking;database
KW - database;data
KW - databases;
KW - databases;visual
KW - indexing;inference
KW - indexing;linear
KW - integrity;database
KW - mechanisms;linear
KW - probabilistic
KW - problem;spatial
KW - problems;temporal
KW - processing;search
KW - program;optimistic
KW - programming;probability;query
KW - query;probabilistic
KW - reasoning;search
KW - selection
KW - selection;consistency
KW - space;cautious
KW - temporal
AB - Spatial probabilistic temporal (SPOT) databases are a paradigm for reasoning with probabilistic statements about where a vehicle may be now or in the future. They express statements of the form "Object O is in spatial region R at some time t with some probability in the interval [L,U]." Past work on SPOT databases has developed selection operators based on selecting SPOT atoms that are entailed by the SPOT database-we call this "cautious" selection. In this paper, we study several problems. First, we note that the runtime of consistency checking and cautious selection algorithms in past work is influenced greatly by the granularity of the underlying Cartesian space. In this paper, we first introduce the notion of "optimistic" selection, where we are interested in returning all SPOT atoms in a database that are consistent with respect to a query, rather than having an entailment relationship. We then develop an approach to scaling SPOT databases that has three main contributions: 1) We develop methods to eliminate variables from the linear programs used in past work, thus greatly reducing the size of the linear programs used-the resulting advances apply to consistency checking, optimistic selection, and cautious selection. 2) We develop a host of theorems to show how we can prune the search space when we are interested in optimistic selection. 3) We use the above contributions to build an efficient index to execute optimistic selection queries over SPOT databases. Our approach is superior to past work in two major respects: First, it makes fewer assumptions than all past works on this topic except that in. Second, our experiments, which are based on real-world data about ship movements, show that our algorithms are much more efficient than those in.
VL - 21
SN - 1041-4347
CP - 1
M3 - 10.1109/TKDE.2008.93
ER -
TY - CONF
T1 - Classifying Computer Generated Charts
T2 - Content-Based Multimedia Indexing, 2007. CBMI '07. International Workshop on
Y1 - 2007
A1 - Prasad,V. S.N
A1 - Siddiquie,B.
A1 - Golbeck,J.
A1 - Davis, Larry S.
KW - algorithm;scale
KW - analysis;visual
KW - classification;image
KW - database;image
KW - databases;
KW - feature
KW - Internet;bar-chart;curve-plot;image
KW - invariant
KW - match
KW - matching;image
KW - relationship;surface-plot;Internet;image
KW - representation;image
KW - segmentation;pie-chart;pyramid
KW - segmentation;statistical
KW - transform;scatter-plot;spatial
AB - We present an approach for classifying images of charts based on the shape and spatial relationships of their primitives. Five categories are considered: bar-charts, curve-plots, pie-charts, scatter-plots and surface-plots. We introduce two novel features to represent the structural information based on (a) region segmentation and (b) curve saliency. The local shape is characterized using the Histograms of Oriented Gradients (HOG) and the Scale Invariant Feature Transform (SIFT) descriptors. Each image is represented by sets of feature vectors of each modality. The similarity between two images is measured by the overlap in the distribution of the features -measured using the Pyramid Match algorithm. A test image is classified based on its similarity with training images from the categories. The approach is tested with a database of images collected from the Internet.
JA - Content-Based Multimedia Indexing, 2007. CBMI '07. International Workshop on
M3 - 10.1109/CBMI.2007.385396
ER -
TY - CONF
T1 - Component-based Data Layout for Efficient Slicing of Very Large Multidimensional Volumetric Data
T2 - Scientific and Statistical Database Management, 2007. SSBDM '07. 19th International Conference on
Y1 - 2007
A1 - Kim,Jusub
A1 - JaJa, Joseph F.
KW - axis-aligned
KW - cache
KW - curves;very
KW - data
KW - data;data
KW - databases;
KW - handling;query
KW - large
KW - layout;data
KW - memory
KW - multidimensional
KW - processing;very
KW - queries;space-filling
KW - size;component-based
KW - slicing
KW - slicing;out-of-core
KW - volumetric
AB - In this paper, we introduce a new efficient data layout scheme to efficiently handle out-of-core axis-aligned slicing queries of very large multidimensional volumetric data. Slicing is a very useful dimension reduction tool that removes or reduces occlusion problems in visualizing 3D/4D volumetric data sets and that enables fast visual exploration of such data sets. We show that the data layouts based on typical space-filling curves are not optimal for the out-of-core slicing queries and present a novel component-based data layout scheme for a specialized problem domain, in which it is only required to provide fast slicing at every k-th value, for any k gt; 1. Our component-based data layout scheme provides much faster processing time for any axis-aligned slicing direction at every k-th value, k gt; 1, requiring less cache memory size and without any replication of data. In addition, the data layout can be generalized to any high dimension.
JA - Scientific and Statistical Database Management, 2007. SSBDM '07. 19th International Conference on
M3 - 10.1109/SSDBM.2007.7
ER -
TY - CONF
T1 - Human Identification using Gait and Face
T2 - Computer Vision and Pattern Recognition, 2007. CVPR '07. IEEE Conference on
Y1 - 2007
A1 - Chellapa, Rama
A1 - Roy-Chowdhury, A.K.
A1 - Kale, A.
KW - algorithm;visual-hull
KW - analysis;image
KW - approach;cameras;face
KW - approximation;probabilistic
KW - database;camera;face
KW - databases;
KW - fusion;face
KW - fusion;human
KW - fusion;probability;video
KW - Gait
KW - identification;planar
KW - NIST
KW - processing;visual
KW - recognition
KW - recognition;gait
KW - recognition;view-invariant
KW - signal
AB - In general the visual-hull approach for performing integrated face and gait recognition requires at least two cameras. In this paper we present experimental results for fusion of face and gait for the single camera case. We considered the NIST database which contains outdoor face and gait data for 30 subjects. In the NIST database, subjects walk along an inverted Sigma pattern. In (A. Kale, et al., 2003), we presented a view-invariant gait recognition algorithm for the single camera case along with some experimental evaluations. In this chapter we present the results of our view-invariant gait recognition algorithm in (A. Kale, et al., 2003) on the NIST database. The algorithm is based on the planar approximation of the person which is valid when the person walks far away from the camera. In (S. Zhou et al., 2003), an algorithm for probabilistic recognition of human faces from video was proposed and the results were demonstrated on the NIST database. Details of these methods can be found in the respective papers. We give an outline of the fusion strategy here.
JA - Computer Vision and Pattern Recognition, 2007. CVPR '07. IEEE Conference on
M3 - 10.1109/CVPR.2007.383523
ER -
TY - CONF
T1 - Indexing Methods for Similarity Searching
T2 - Current Trends in Computer Science, 2007. ENC 2007. Eighth Mexican International Conference on
Y1 - 2007
A1 - Samet, Hanan
KW - content-based
KW - database;indexing
KW - database;pattern
KW - databases;
KW - indexing;multimedia
KW - methods;multimedia
KW - recognition;similarity
KW - retrieval;data
KW - searching;data
KW - structures;database
KW - structures;image
AB - An overview is given of the various techniques and issues involved in providing indexing support for similarity searching. Similarity searching is a crucial part of retrieval in multimedia databases used for applications such as pattern recognition, image databases, and content-based retrieval. It involves finding objects in a data set S that are similar to a query object q based on some distance measure d which is usually a distance metric. The search process is usually achieved by means of nearest neighbor finding. Existing methods for handling similarity search in this setting fall into one of two classes. The first is based on mapping to a vector space. The vector space is usually of high dimension which requires special handling due to the fact indexing methods do not discriminate well in such spaces. In particular, the query regions often overlap all of the blocks that result from the decomposition of the underlying space. This has led to some special solutions that make use of a sequential scan. An alternative is to use dimension reduction to find a mapping from a high-dimensional space into a low-dimensional space by finding the most discriminating dimensions and then index the data using one of a number of different data structures such as k-d trees, R-trees, quadtrees, etc. The second directly indexes the objects based on distances making use of data structures such as the vp-tree, M-tree, etc.
JA - Current Trends in Computer Science, 2007. ENC 2007. Eighth Mexican International Conference on
M3 - 10.1109/ENC.2007.9
ER -
TY - CONF
T1 - Indexing Point Triples Via Triangle Geometry
T2 - Data Engineering, 2007. ICDE 2007. IEEE 23rd International Conference on
Y1 - 2007
A1 - Cranston,C.B.
A1 - Samet, Hanan
KW - database
KW - databases;
KW - dimension;spatial
KW - geometry;database
KW - hyperdimensional
KW - index
KW - index;single
KW - index;triangle
KW - indexing;query
KW - linear
KW - point;k-fold
KW - processing;visual
KW - relationships;structured
KW - rotational
KW - search;indexing
KW - space;image
KW - symmetry;point-based
AB - Database search for images containing icons with specific mutual spatial relationships can be facilitated by an appropriately structured index. For the case of images containing subsets each of which consist of three icons, the one-to-one correspondence between (distinct) point triples and triangles allows the use of such triangle attributes as position, size, orientation, and "shape" in constructing a point-based index, in which each triangle maps to a single point in a resulting hyperdimensional index space. Size (based on the triangle perimeter) can be represented by a single linear dimension. The abstract "shape" of a triangle induces a space that is inherently two-dimensional, and a number of alternative definitions of a basis for this space are examined. Within a plane, orientation reduces to rotation, and (after assignment of a reference direction for the triangle) can be represented by a single, spatially closed dimension. However, assignment of a reference direction for triangles possessing a k-fold rotational symmetry presents a significant challenge. Methods are described for characterizing shape and orientation of triangles, and for mapping these attributes onto a set of linear axes to form a combined index. The shape attribute is independent of size, orientation, and position, and the characterization of shape and orientation is stable with respect to small variations in the indexed triangles.
JA - Data Engineering, 2007. ICDE 2007. IEEE 23rd International Conference on
M3 - 10.1109/ICDE.2007.367939
ER -
TY - CONF
T1 - Mitigating risk of data loss in preservation environments
T2 - Mass Storage Systems and Technologies, 2005. Proceedings. 22nd IEEE / 13th NASA Goddard Conference on
Y1 - 2005
A1 - Moore,R.W.
A1 - JaJa, Joseph F.
A1 - Chadduck,R.
KW - archives;
KW - authentication;
KW - authenticity;
KW - computing;
KW - data
KW - databases;
KW - digital
KW - distributed
KW - environment;
KW - Grid
KW - integrity;
KW - management;
KW - message
KW - objects;
KW - persistent
KW - preservation
KW - record
KW - risk
KW - storage
AB - Preservation environments manage digital records for time periods that are much longer than that of a single vendor product. A primary requirement is the preservation of the authenticity and integrity of the digital records while simultaneously minimizing the cost of long-term storage, as the data is migrated onto successive generations of technology. The emergence of low-cost storage hardware has made it possible to implement innovative software systems that minimize risk of data loss and preserve authenticity and integrity. This paper describes software mechanisms in use in current persistent archives and presents an example based upon the NARA research prototype persistent archive.
JA - Mass Storage Systems and Technologies, 2005. Proceedings. 22nd IEEE / 13th NASA Goddard Conference on
M3 - 10.1109/MSST.2005.20
ER -
TY - CONF
T1 - RDF aggregate queries and views
T2 - Data Engineering, 2005. ICDE 2005. Proceedings. 21st International Conference on
Y1 - 2005
A1 - Hung,E.
A1 - Deng,Yu
A1 - V.S. Subrahmanian
KW - aggregate
KW - databases;
KW - DBMS;
KW - description
KW - framework;
KW - languages;
KW - Maintenance
KW - methods;
KW - processing;
KW - queries;
KW - query
KW - RDF
KW - relational
KW - resource
KW - standard;
KW - standards;
KW - view
KW - Web
AB - Resource description framework (RDF) is a rapidly expanding Web standard. RDF databases attempt to track the massive amounts of Web data and services available. In this paper, we study the problem of aggregate queries. We develop an algorithm to compute answers to aggregate queries over RDF databases and algorithms to maintain views involving those aggregates. Though RDF data can be stored in a standard relational DBMS (and hence we can execute standard relational aggregate queries and view maintenance methods on them), we show experimentally that our algorithms that operate directly on the RDF representation exhibit significantly superior performance.
JA - Data Engineering, 2005. ICDE 2005. Proceedings. 21st International Conference on
M3 - 10.1109/ICDE.2005.121
ER -
TY - CONF
T1 - Using the inner-distance for classification of articulated shapes
T2 - Computer Vision and Pattern Recognition, 2005. CVPR 2005. IEEE Computer Society Conference on
Y1 - 2005
A1 - Ling,H.
A1 - Jacobs, David W.
KW - articulated
KW - CE-Shape-1
KW - classification;
KW - database;
KW - databases;
KW - dataset;
KW - descriptor;
KW - dynamic
KW - human
KW - image
KW - inner-distance;
KW - Kimia
KW - landmark
KW - leaf
KW - matching;
KW - MOTION
KW - MPEG7
KW - points;
KW - programming;
KW - SHAPE
KW - silhouette
KW - silhouette;
KW - Swedish
KW - visual
AB - We propose using the inner-distance between landmark points to build shape descriptors. The inner-distance is defined as the length of the shortest path between landmark points within the shape silhouette. We show that the inner-distance is articulation insensitive and more effective at capturing complex shapes with part structures than Euclidean distance. To demonstrate this idea, it is used to build a new shape descriptor based on shape contexts. After that, we design a dynamic programming based method for shape matching and comparison. We have tested our approach on a variety of shape databases including an articulated shape dataset, MPEG7 CE-Shape-1, Kimia silhouettes, a Swedish leaf database and a human motion silhouette dataset. In all the experiments, our method demonstrates effective performance compared with other algorithms.
JA - Computer Vision and Pattern Recognition, 2005. CVPR 2005. IEEE Computer Society Conference on
VL - 2
M3 - 10.1109/CVPR.2005.362
ER -
TY - CONF
T1 - Adaptive replication in peer-to-peer systems
T2 - Distributed Computing Systems, 2004. Proceedings. 24th International Conference on
Y1 - 2004
A1 - Gopalakrishnan,V.
A1 - Silaghi,B.
A1 - Bhattacharjee, Bobby
A1 - Keleher,P.
KW - adaptive
KW - allocation;
KW - data
KW - databases;
KW - decentralized
KW - delivery
KW - distributed
KW - LAR
KW - low-latency
KW - peer-to-peer
KW - processing;
KW - protocol;
KW - replicated
KW - replication
KW - resource
KW - strategies;
KW - structured
KW - system-neutral
KW - system;
KW - systems;
AB - Peer-to-peer systems can be used to form a low-latency decentralized data delivery system. Structured peer-to-peer systems provide both low latency and excellent load balance with uniform query and data distributions. Under the more common skewed access distributions, however, individual nodes are easily overloaded, resulting in poor global performance and lost messages. This paper describes a lightweight, adaptive, and system-neutral replication protocol, called LAR, that maintains low access latencies and good load balance even under highly skewed demand. We apply LAR to Chord and show that it has lower overhead and better performance than existing replication strategies.
JA - Distributed Computing Systems, 2004. Proceedings. 24th International Conference on
M3 - 10.1109/ICDCS.2004.1281601
ER -
TY - CONF
T1 - Facial similarity across age, disguise, illumination and pose
T2 - Image Processing, 2004. ICIP '04. 2004 International Conference on
Y1 - 2004
A1 - Ramanathan,N.
A1 - Chellapa, Rama
A1 - Roy Chowdhury, A.K.
KW - Aging
KW - database
KW - databases;
KW - disguise;
KW - effect;
KW - Expression
KW - Face
KW - half-face;
KW - illumination;
KW - image
KW - lighting;
KW - pose
KW - recognition
KW - recognition;
KW - retrieval;
KW - system;
KW - variation;
KW - visual
AB - Illumination, pose variations, disguises, aging effects and expression variations are some of the key factors that affect the performance of face recognition systems. Face recognition systems have always been studied from a recognition perspective. Our emphasis is on deriving a measure of similarity between faces. The similarity measure provides insights into the role each of the above mentioned variations play in affecting the performance of face recognition systems. In the process of computing the similarity measure between faces, we suggest a framework to compensate for pose variations and introduce the notion of 'half-faces' to circumvent the problem of non-uniform illumination. We used the similarity measure to retrieve similar faces from a database containing multiple images of individuals. Moreover, we devised experiments to study the effect age plays in affecting facial similarity. In conclusion, the similarity measure helps in studying the significance facial features play in affecting the performance of face recognition systems.
JA - Image Processing, 2004. ICIP '04. 2004 International Conference on
VL - 3
M3 - 10.1109/ICIP.2004.1421474
ER -
TY - CONF
T1 - Multiple-exemplar discriminant analysis for face recognition
T2 - Pattern Recognition, 2004. ICPR 2004. Proceedings of the 17th International Conference on
Y1 - 2004
A1 - Zhou,S. K
A1 - Chellapa, Rama
KW - analysis;
KW - database;
KW - databases;
KW - discriminant
KW - Face
KW - FERET
KW - multiple-exemplar
KW - recognition;
KW - visual
AB - Face recognition is characteristically different from regular pattern recognition and, therefore, requires a different discriminant analysis other than linear discriminant analysis(LDA). LDA is a single-exemplar method in the sense that each class during classification is represented by a single exemplar, i.e., the sample mean of the class. We present a multiple-exemplar discriminant analysis (MEDA) where each class is represented using several exemplars or even the whole available sample set. The proposed approach produces improved classification results when tested on a subset of FERET database where LDA is ineffective.
JA - Pattern Recognition, 2004. ICPR 2004. Proceedings of the 17th International Conference on
VL - 4
M3 - 10.1109/ICPR.2004.1333736
ER -
TY - CONF
T1 - Strategies for exploring large scale data
T2 - Parallel Architectures, Algorithms and Networks, 2004. Proceedings. 7th International Symposium on
Y1 - 2004
A1 - JaJa, Joseph F.
KW - algorithms;
KW - association;
KW - asymptotic
KW - bounds;
KW - business
KW - data
KW - data;
KW - database
KW - databases;
KW - demographic
KW - discovery;
KW - Indexing
KW - indexing;
KW - information
KW - knowledge
KW - large
KW - linear
KW - mining;
KW - multidimensional
KW - objects;
KW - optimal
KW - Parallel
KW - pattern
KW - processing;
KW - query
KW - querying;
KW - range
KW - scale
KW - scientific
KW - search
KW - serial
KW - series
KW - series;
KW - simulation
KW - space
KW - structure;
KW - structures;
KW - techniques;
KW - temporal
KW - TIME
KW - value
KW - very
KW - window;
AB - We consider the problem of querying large scale multidimensional time series data to discover events of interest, test and validate hypotheses, or to associate temporal patterns with specific events. This type of data currently dominates most other types of available data, and will very likely become even more prevalent in the future given the current trends in collecting time series of business, scientific, demographic, and simulation data. The ability to explore such collections interactively, even at a coarse level, will be critical in discovering the information and knowledge embedded in such collections. We develop indexing techniques and search algorithms to efficiently handle temporal range value querying of multidimensional time series data. Our indexing uses linear space data structures that enable the handling of queries in I/O time that is essentially the same as that of handling a single time slice, assuming the availability of a logarithmic number of processors as a function of the temporal window. A data structure with provably almost optimal asymptotic bounds is also presented for the case when the number of multidimensional objects is relatively small. These techniques improve significantly over standard techniques for either serial or parallel processing, and are evaluated by extensive experimental results that confirm their superior performance.
JA - Parallel Architectures, Algorithms and Networks, 2004. Proceedings. 7th International Symposium on
M3 - 10.1109/ISPAN.2004.1300447
ER -
TY - JOUR
T1 - Deno: a decentralized, peer-to-peer object-replication system for weakly connected environments
JF - Computers, IEEE Transactions on
Y1 - 2003
A1 - Cetintemel,U.
A1 - Keleher,P. J
A1 - Bhattacharjee, Bobby
A1 - Franklin,M.J.
KW - actions;
KW - connected
KW - consistency
KW - data
KW - data;
KW - databases;
KW - decentralized
KW - Deno;
KW - distributed
KW - environments;
KW - epidemic
KW - group
KW - levels;
KW - Linux;
KW - malicious
KW - management;
KW - membership;
KW - network
KW - object
KW - object-replication
KW - of
KW - operating
KW - peer-to-peer
KW - protocols;
KW - replicated
KW - replication;
KW - Security
KW - security;
KW - synchronisation;
KW - system;
KW - systems;
KW - topology;
KW - Unix;
KW - voting;
KW - weakly
KW - weighted
KW - Win32;
AB - This paper presents the design, implementation, and evaluation of the replication framework of Deno, a decentralized, peer-to-peer object-replication system targeted for weakly connected environments. Deno uses weighted voting for availability and pair-wise, epidemic information flow for flexibility. This combination allows the protocols to operate with less than full connectivity, to easily adapt to changes in group membership, and to make few assumptions about the underlying network topology. We present two versions of Deno's protocol that differ in the consistency levels they support. We also propose security extensions to handle a class of malicious actions that involve misrepresentation of protocol information. Deno has been implemented and runs on top of Linux and Win32 platforms. We use the Deno prototype to characterize the performance of the Deno protocols and extensions. Our study reveals several interesting results that provide fundamental insight into the benefits of decentralization and the mechanics of epidemic protocols.
VL - 52
SN - 0018-9340
CP - 7
M3 - 10.1109/TC.2003.1214342
ER -
TY - JOUR
T1 - Properties of embedding methods for similarity searching in metric spaces
JF - Pattern Analysis and Machine Intelligence, IEEE Transactions on
Y1 - 2003
A1 - Hjaltason,G. R
A1 - Samet, Hanan
KW - complex
KW - contractiveness;
KW - data
KW - databases;
KW - decomposition;
KW - dimension
KW - distance
KW - distortion;
KW - DNA
KW - documents;
KW - EMBEDDING
KW - embeddings;
KW - Euclidean
KW - evaluations;
KW - FastMap;
KW - images;
KW - Lipschitz
KW - methods;
KW - metric
KW - MetricMap;
KW - multimedia
KW - processing;
KW - query
KW - reduction
KW - search;
KW - searching;
KW - sequences;
KW - similarity
KW - singular
KW - spaces;
KW - SparseMap;
KW - types;
KW - value
AB - Complex data types-such as images, documents, DNA sequences, etc.-are becoming increasingly important in modern database applications. A typical query in many of these applications seeks to find objects that are similar to some target object, where (dis)similarity is defined by some distance function. Often, the cost of evaluating the distance between two objects is very high. Thus, the number of distance evaluations should be kept at a minimum, while (ideally) maintaining the quality of the result. One way to approach this goal is to embed the data objects in a vector space so that the distances of the embedded objects approximates the actual distances. Thus, queries can be performed (for the most part) on the embedded objects. We are especially interested in examining the issue of whether or not the embedding methods will ensure that no relevant objects are left out. Particular attention is paid to the SparseMap, FastMap, and MetricMap embedding methods. SparseMap is a variant of Lipschitz embeddings, while FastMap and MetricMap are inspired by dimension reduction methods for Euclidean spaces. We show that, in general, none of these embedding methods guarantee that queries on the embedded objects have no false dismissals, while also demonstrating the limited cases in which the guarantee does hold. Moreover, we describe a variant of SparseMap that allows queries with no false dismissals. In addition, we show that with FastMap and MetricMap, the distances of the embedded objects can be much greater than the actual distances. This makes it impossible (or at least impractical) to modify FastMap and MetricMap to guarantee no false dismissals.
VL - 25
SN - 0162-8828
CP - 5
M3 - 10.1109/TPAMI.2003.1195989
ER -
TY - CONF
T1 - PXML: a probabilistic semistructured data model and algebra
T2 - Data Engineering, 2003. Proceedings. 19th International Conference on
Y1 - 2003
A1 - Hung,E.
A1 - Getoor, Lise
A1 - V.S. Subrahmanian
KW - algebra;
KW - data
KW - databases;
KW - instances;
KW - model;
KW - models;
KW - probabilistic
KW - processing;
KW - PXML;
KW - query
KW - relational
KW - semistructured
KW - structures;
KW - tree
KW - XML;
AB - Despite the recent proliferation of work on semistructured data models, there has been little work to date on supporting uncertainty in these models. We propose a model for probabilistic semistructured data (PSD). The advantage of our approach is that it supports a flexible representation that allows the specification of a wide class of distributions over semistructured instances. We provide two semantics for the model and show that the semantics are probabilistically coherent. Next, we develop an extension of the relational algebra to handle probabilistic semistructured data and describe efficient algorithms for answering queries that use this algebra. Finally, we present experimental results showing the efficiency of our algorithms.
JA - Data Engineering, 2003. Proceedings. 19th International Conference on
M3 - 10.1109/ICDE.2003.1260814
ER -
TY - CONF
T1 - Rank constrained recognition under unknown illuminations
T2 - Analysis and Modeling of Faces and Gestures, 2003. AMFG 2003. IEEE International Workshop on
Y1 - 2003
A1 - Zhou, S.
A1 - Chellapa, Rama
KW - albedo
KW - approach;
KW - constrained
KW - database;
KW - databases;
KW - decomposition;
KW - factorization
KW - field;
KW - illumination
KW - image;
KW - information;
KW - Lambertian
KW - lighting;
KW - model;
KW - object
KW - object-specific
KW - PIE
KW - rank
KW - recognition;
KW - reflectance
KW - samples;
KW - singular
KW - three-dimensional
KW - two-dimensional
KW - value
KW - variations;
KW - visual
AB - Recognition under illumination variations is a challenging problem. The key is to successfully separate the illumination source from the observed appearance. Once separated, what remains is invariant to illuminant and appropriate for recognition. Most current efforts employ a Lambertian reflectance model with varying albedo field ignoring both attached and cast shadows, but restrict themselves by using object-specific samples, which undesirably deprives them of recognizing new objects not in the training samples. Using rank constraints on the albedo and the surface normal, we accomplish illumination separation in a more general setting, e.g., with class-specific samples via a factorization approach. In addition, we handle shadows (both attached and cast ones) by treating them as missing values, and resolve the ambiguities in the factorization method by enforcing integrability. As far as recognition is concerned, a bootstrap set which is just a collection of two-dimensional image observations can be utilized to avoid the explicit requirement that three-dimensional information be available. Our approaches produce good recognition results as shown in our experiments using the PIE database.
JA - Analysis and Modeling of Faces and Gestures, 2003. AMFG 2003. IEEE International Workshop on
M3 - 10.1109/AMFG.2003.1240818
ER -
TY - JOUR
T1 - Temporal probabilistic object bases
JF - Knowledge and Data Engineering, IEEE Transactions on
Y1 - 2003
A1 - Biazzo,V.
A1 - Giugno,R.
A1 - Lukasiewicz,T.
A1 - V.S. Subrahmanian
KW - algebra;
KW - algebraic
KW - bases;
KW - constraints;
KW - data
KW - database
KW - database;
KW - databases;
KW - distribution
KW - explicit
KW - functions;
KW - handling;
KW - implicit
KW - instances;
KW - integrity;
KW - intervals;
KW - management;
KW - model;
KW - models;
KW - object
KW - object-oriented
KW - operations;
KW - probabilistic
KW - probability
KW - probability;
KW - relational
KW - temporal
KW - theory;
KW - Uncertainty
KW - uncertainty;
AB - There are numerous applications where we have to deal with temporal uncertainty associated with objects. The ability to automatically store and manipulate time, probabilities, and objects is important. We propose a data model and algebra for temporal probabilistic object bases (TPOBs), which allows us to specify the probability with which an event occurs at a given time point. In explicit TPOB-instances, the sets of time points along with their probability intervals are explicitly enumerated. In implicit TPOB-instances, sets of time points are expressed by constraints and their probability intervals by probability distribution functions. Thus, implicit object base instances are succinct representations of explicit ones; they allow for an efficient implementation of algebraic operations, while their explicit counterparts make defining algebraic operations easy. We extend the relational algebra to both explicit and implicit instances and prove that the operations on implicit instances correctly implement their counterpart on explicit instances.
VL - 15
SN - 1041-4347
CP - 4
M3 - 10.1109/TKDE.2003.1209009
ER -
TY - CONF
T1 - Content-based image retrieval using Fourier descriptors on a logo database
T2 - Pattern Recognition, 2002. Proceedings. 16th International Conference on
Y1 - 2002
A1 - Folkers,A.
A1 - Samet, Hanan
KW - abstraction;
KW - analysis;
KW - constraints;
KW - content-based
KW - contour
KW - database
KW - database;
KW - databases;
KW - descriptors;
KW - detection;
KW - edge
KW - Fourier
KW - image
KW - logos;
KW - pictorial
KW - processing;
KW - query
KW - retrieval;
KW - SHAPE
KW - spatial
KW - specification;
KW - theory;
KW - visual
AB - A system that enables the pictorial specification of queries in an image database is described. The queries are comprised of rectangle, polygon, ellipse, and B-spline shapes. The queries specify which shapes should appear in the target image as well as spatial constraints on the distance between them and their relative position. The retrieval process makes use of an abstraction of the contour of the shape which is invariant against translation, scale, rotation, and starting point, that is based on the use of Fourier descriptors. These abstractions are used in a system to locate logos in an image database. The utility of this approach is illustrated using some sample queries.
JA - Pattern Recognition, 2002. Proceedings. 16th International Conference on
VL - 3
M3 - 10.1109/ICPR.2002.1047991
ER -
TY - CONF
T1 - Efficient techniques for range search queries on earth science data
T2 - Scientific and Statistical Database Management, 2002. Proceedings. 14th International Conference on
Y1 - 2002
A1 - Shi,Qingmin
A1 - JaJa, Joseph F.
KW - based
KW - computing;
KW - content
KW - data
KW - data;
KW - databases;
KW - Earth
KW - factors;
KW - large
KW - mining
KW - mining;
KW - natural
KW - processing;
KW - queries;
KW - query
KW - range
KW - raster
KW - retrieval;
KW - scale
KW - Science
KW - sciences
KW - search
KW - spatial
KW - structures;
KW - tasks;
KW - temporal
KW - tree
KW - tree-of-regions;
KW - visual
AB - We consider the problem of organizing large scale earth science raster data to efficiently handle queries for identifying regions whose parameters fall within certain range values specified by the queries. This problem seems to be critical to enabling basic data mining tasks such as determining associations between physical phenomena and spatial factors, detecting changes and trends, and content based retrieval. We assume that the input is too large to fit in internal memory and hence focus on data structures and algorithms that minimize the I/O bounds. A new data structure, called a tree-of-regions (ToR), is introduced and involves a combination of an R-tree and efficient representation of regions. It is shown that such a data structure enables the handling of range queries in an optimal I/O time, under certain reasonable assumptions. We also show that updates to the ToR can be handled efficiently. Experimental results for a variety of multi-valued earth science data illustrate the fast execution times of a wide range of queries, as predicted by our theoretical analysis.
JA - Scientific and Statistical Database Management, 2002. Proceedings. 14th International Conference on
M3 - 10.1109/SSDM.2002.1029714
ER -
TY - JOUR
T1 - Presentation planning for distributed VoD systems
JF - Knowledge and Data Engineering, IEEE Transactions on
Y1 - 2002
A1 - Hwang,Eenjun
A1 - Prabhakaran,B.
A1 - V.S. Subrahmanian
KW - Computer
KW - computing;
KW - databases;
KW - demand;
KW - distributed
KW - local
KW - multimedia
KW - network;
KW - on
KW - optimal
KW - plan;
KW - plans;
KW - presentation
KW - presentation;
KW - server;
KW - servers;
KW - video
KW - video-on-demand;
KW - VoD;
AB - A distributed video-on-demand (VoD) system is one where a collection of video data is located at dispersed sites across a computer network. In a single site environment, a local video server retrieves video data from its local storage device. However, in distributed VoD systems, when a customer requests a movie from the local server, the server may need to interact with other servers located across the network. In this paper, we present different types of presentation plans that a local server can construct in order to satisfy a customer request. Informally speaking, a presentation plan is a temporally synchronized sequence of steps that the local server must perform in order to present the requested movie to the customer. This involves obtaining commitments from other video servers, obtaining commitments from the network service provider, as well as making commitments of local resources, while keeping within the limitations of available bandwidth, available buffer, and customer data consumption rates. Furthermore, in order to evaluate the quality of a presentation plan, we introduce two measures of optimality for presentation plans: minimizing wait time for a customer and minimizing access bandwidth which, informally speaking, specifies how much network/disk bandwidth is used. We develop algorithms to compute three different optimal presentation plans that work at a block level, or at a segment level, or with a hybrid mix of the two, and compare their performance through simulation experiments. We have also mathematically proven effects of increased buffer or bandwidth and data replications for presentation plans which had previously been verified experimentally in the literature.
VL - 14
SN - 1041-4347
CP - 5
M3 - 10.1109/TKDE.2002.1033774
ER -
TY - CONF
T1 - Condensing image databases when retrieval is based on non-metric distances
T2 - Computer Vision, 1998. Sixth International Conference on
Y1 - 1998
A1 - Jacobs, David W.
A1 - Weinshall,D.
A1 - Gdalyahu,Y.
KW - appearance-based
KW - databases;
KW - databases;image
KW - dataspaces;pattern
KW - MATCHING
KW - methods;non-metric
KW - processing;visual
KW - recognition;query
KW - systems;image
KW - vision;classification
AB - One of the key problems in appearance-based vision is understanding how to use a set of labeled images to classify new images. Classification systems that can model human performance, or that use robust image matching methods, often make use of similarity judgments that are non-metric but when the triangle inequality is not obeyed, most existing pattern recognition techniques are not applicable. We note that exemplar-based (or nearest-neighbor) methods can be applied naturally when using a wide class of non-metric similarity functions. The key issue, however, is to find methods for choosing good representatives of a class that accurately characterize it. We note that existing condensing techniques for finding class representatives are ill-suited to deal with non-metric dataspaces. We then focus on developing techniques for solving this problem, emphasizing two points: First, we show that the distance between two images is not a good measure of how well one image can represent another in non-metric spaces. Instead, we use the vector correlation between the distances from each image to other previously seen images. Second, we show that in non-metric spaces, boundary points are less significant for capturing the structure of a class than they are in Euclidean spaces. We suggest that atypical points may be more important in describing classes. We demonstrate the importance of these ideas to learning that generalizes from experience by improving performance using both synthetic and real images
JA - Computer Vision, 1998. Sixth International Conference on
M3 - 10.1109/ICCV.1998.710778
ER -
TY - CONF
T1 - The role of compressed document images in transmission and retrieval
T2 - Multimedia Signal Processing, 1997., IEEE First Workshop on
Y1 - 1997
A1 - Kia,O.
A1 - David Doermann
KW - applications;progressive
KW - coding;multimedia
KW - component
KW - compression;document
KW - compression;image
KW - computing;query
KW - content;lossy
KW - databases;
KW - distortion
KW - experiment;structural
KW - hierarchy;symbols;visual
KW - image
KW - level
KW - objects;data
KW - processing;image
KW - processing;visual
KW - retrieval;image
KW - structure;document
KW - transmission;information
KW - transmission;multimedia;network
KW - transmission;rate
AB - Document images belong to a unique class of images where the information content is contained in the language represented by a series of symbols on the page, rather than in the visual objects themselves. For this reason, it is essential to preserve the fidelity of individual components when considering methods of compression. Likewise the component level structure should be a prime consideration when ordering information for lossy or progressive transmission. We refine our work on document image compression as it applies to transmission and retrieval. We first overview the basic compression scheme, then describe a structural hierarchy which provides desirable properties for transmission. We present the results of a rate distortion experiment and discuss the implications for network applications
JA - Multimedia Signal Processing, 1997., IEEE First Workshop on
M3 - 10.1109/MMSP.1997.602657
ER -