Scene Classification Using a Hybrid Generative/Discriminative Approach

We investigate whether dimensionality reduction using a latent generative model is beneficial for the task of weakly supervised scene classification. In detail, we are given a set of labeled images of scenes (for example, coast, forest, city, river, etc.), and our objective is to classify a new image into one of these categories. Our approach consists of first discovering latent ";topics"; using probabilistic Latent Semantic Analysis (pLSA), a generative model from the statistical text literature here applied to a bag of visual words representation for each image, and subsequently, training a multiway classifier on the topic distribution vector for each image. We compare this approach to that of representing each image by a bag of visual words vector directly and training a multiway classifier on these vectors. To this end, we introduce a novel vocabulary using dense color SIFT descriptors and then investigate the classification performance under changes in the size of the visual vocabulary, the number of latent topics learned, and the type of discriminative classifier used (k-nearest neighbor or SVM). We achieve superior classification performance to recent publications that have used a bag of visual word representation, in all cases, using the authors' own data sets and testing protocols. We also investigate the gain in adding spatial information. We show applications to image retrieval with relevance feedback and to scene classification in videos

Intersecciones entre visión artificial y mirada artística. Semantic Intersections between Artificial and Artistic Vision.

En el presente artículo se ha desarrollado un sistema capaz de categorizar de forma automática la base de datos de imágenes que sirven de punto de partida para la ideación y diseño en la producción artística del escultor M. Planas. La metodología utilizada está basada en características locales. Para la construcción de un vocabulario visual se sigue un procedimiento análogo al que se utiliza en el análisis automático de textos (modelo 'Bag-of-Words'-BOW) y en el ámbito de las imágenes nos referiremos a representaciones 'Bag-of-Visual Terms' (BOV). En este enfoque se analizan las imágenes como un conjunto de regiones, describiendo solamente su apariencia e ignorando su estructura espacial. Para superar los inconvenientes de polisemia y sinonimia que lleva asociados esta metodología, se utiliza el análisis probabilístico de aspectos latentes (PLSA) que detecta aspectos subyacentes en las imágenes, patrones formales. Los resultados obtenidos son prometedores y, además de la utilidad intrínseca de la categorización automática de imágenes, este método puede proporcionar al artista un punto de vista auxiliar muy interesante.

Modeling and Classifying Breast Tissue Density in Mammograms

We present a new approach to model and classify breast parenchymal tissue. Given a mammogram, first, we will discover the distribution of the different tissue densities in an unsupervised manner, and second, we will use this tissue distribution to perform the classification. We achieve this using a classifier based on local descriptors and probabilistic Latent Semantic Analysis (pLSA), a generative model from the statistical text literature. We studied the influence of different descriptors like texture and SIFT features at the classification stage showing that textons outperform SIFT in all cases. Moreover we demonstrate that pLSA automatically extracts meaningful latent aspects generating a compact tissue representation based on their densities, useful for discriminating on mammogram classification. We show the results of tissue classification over the MIAS and DDSM datasets. We compare our method with approaches that classified these same datasets showing a better performance of our proposal

Modeling and Classifying Breast Tissue Density in Mammograms

We present a new approach to model and classify breast parenchymal tissue. Given a mammogram, first, we will discover the distribution of the different tissue densities in an unsupervised manner, and second, we will use this tissue distribution to perform the classification. We achieve this using a classifier based on local descriptors and probabilistic Latent Semantic Analysis (pLSA), a generative model from the statistical text literature. We studied the influence of different descriptors like texture and SIFT features at the classification stage showing that textons outperform SIFT in all cases. Moreover we demonstrate that pLSA automatically extracts meaningful latent aspects generating a compact tissue representation based on their densities, useful for discriminating on mammogram classification. We show the results of tissue classification over the MIAS and DDSM datasets. We compare our method with approaches that classified these same datasets showing a better performance of our proposal

Scene Classification Using a Hybrid Generative/Discriminative Approach

We investigate whether dimensionality reduction using a latent generative model is beneficial for the task of weakly supervised scene classification. In detail, we are given a set of labeled images of scenes (for example, coast, forest, city, river, etc.), and our objective is to classify a new image into one of these categories. Our approach consists of first discovering latent ";topics"; using probabilistic Latent Semantic Analysis (pLSA), a generative model from the statistical text literature here applied to a bag of visual words representation for each image, and subsequently, training a multiway classifier on the topic distribution vector for each image. We compare this approach to that of representing each image by a bag of visual words vector directly and training a multiway classifier on these vectors. To this end, we introduce a novel vocabulary using dense color SIFT descriptors and then investigate the classification performance under changes in the size of the visual vocabulary, the number of latent topics learned, and the type of discriminative classifier used (k-nearest neighbor or SVM). We achieve superior classification performance to recent publications that have used a bag of visual word representation, in all cases, using the authors' own data sets and testing protocols. We also investigate the gain in adding spatial information. We show applications to image retrieval with relevance feedback and to scene classification in videos

Image classification for a large number of object categories

L'increment de bases de dades que cada vegada contenen imatges més difícils i amb un nombre més elevat de categories, està forçant el desenvolupament de tècniques de representació d'imatges que siguin discriminatives quan es vol treballar amb múltiples classes i d'algorismes que siguin eficients en l'aprenentatge i classificació. Aquesta tesi explora el problema de classificar les imatges segons l'objecte que contenen quan es disposa d'un gran nombre de categories. Primerament s'investiga com un sistema híbrid format per un model generatiu i un model discriminatiu pot beneficiar la tasca de classificació d'imatges on el nivell d'anotació humà sigui mínim. Per aquesta tasca introduïm un nou vocabulari utilitzant una representació densa de descriptors color-SIFT, i desprès s'investiga com els diferents paràmetres afecten la classificació final. Tot seguit es proposa un mètode par tal d'incorporar informació espacial amb el sistema híbrid, mostrant que la informació de context es de gran ajuda per la classificació d'imatges. Desprès introduïm un nou descriptor de forma que representa la imatge segons la seva forma local i la seva forma espacial, tot junt amb un kernel que incorpora aquesta informació espacial en forma piramidal. La forma es representada per un vector compacte obtenint un descriptor molt adequat per ésser utilitzat amb algorismes d'aprenentatge amb kernels. Els experiments realitzats postren que aquesta informació de forma te uns resultats semblants (i a vegades millors) als descriptors basats en aparença. També s'investiga com diferents característiques es poden combinar per ésser utilitzades en la classificació d'imatges i es mostra com el descriptor de forma proposat juntament amb un descriptor d'aparença millora substancialment la classificació. Finalment es descriu un algoritme que detecta les regions d'interès automàticament durant l'entrenament i la classificació. Això proporciona un mètode per inhibir el fons de la imatge i afegeix invariança a la posició dels objectes dins les imatges. S'ensenya que la forma i l'aparença sobre aquesta regió d'interès i utilitzant els classificadors random forests millora la classificació i el temps computacional. Es comparen els postres resultats amb resultats de la literatura utilitzant les mateixes bases de dades que els autors Aixa com els mateixos protocols d'aprenentatge i classificació. Es veu com totes les innovacions introduïdes incrementen la classificació final de les imatges.

Discovering user access pattern based on probabilistic latent factor model

There has been an increased demand for characterizing user access patterns using web mining techniques since the informative knowledge extracted from web server log files can not only offer benefits for web site structure improvement but also for better understanding of user navigational behavior. In this paper, we present a web usage mining method, which utilize web user usage and page linkage information to capture user access pattern based on Probabilistic Latent Semantic Analysis (PLSA) model. A specific probabilistic model analysis algorithm, EM algorithm, is applied to the integrated usage data to infer the latent semantic factors as well as generate user session clusters for revealing user access patterns. Experiments have been conducted on real world data set to validate the effectiveness of the proposed approach. The results have shown that the presented method is capable of characterizing the latent semantic factors and generating user profile in terms of weighted page vectors, which may reflect the common access interest exhibited by users among same session cluster.

A web recommendation technique based on probabilistic latent semantic analysis

Web transaction data between Web visitors and Web functionalities usually convey user task-oriented behavior pattern. Mining such type of click-stream data will lead to capture usage pattern information. Nowadays Web usage mining technique has become one of most widely used methods for Web recommendation, which customizes Web content to user-preferred style. Traditional techniques of Web usage mining, such as Web user session or Web page clustering, association rule and frequent navigational path mining can only discover usage pattern explicitly. They, however, cannot reveal the underlying navigational activities and identify the latent relationships that are associated with the patterns among Web users as well as Web pages. In this work, we propose a Web recommendation framework incorporating Web usage mining technique based on Probabilistic Latent Semantic Analysis (PLSA) model. The main advantages of this method are, not only to discover usage-based access pattern, but also to reveal the underlying latent factor as well. With the discovered user access pattern, we then present user more interested content via collaborative recommendation. To validate the effectiveness of proposed approach, we conduct experiments on real world datasets and make comparisons with some existing traditional techniques. The preliminary experimental results demonstrate the usability of the proposed approach.

Towards user profiling for web recommendation

Collaborative recommendation is one of widely used recommendation systems, which recommend items to visitor on a basis of referring other's preference that is similar to current user. User profiling technique upon Web transaction data is able to capture such informative knowledge of user task or interest. With the discovered usage pattern information, it is likely to recommend Web users more preferred content or customize the Web presentation to visitors via collaborative recommendation. In addition, it is helpful to identify the underlying relationships among Web users, items as well as latent tasks during Web mining period. In this paper, we propose a Web recommendation framework based on user profiling technique. In this approach, we employ Probabilistic Latent Semantic Analysis (PLSA) to model the co-occurrence activities and develop a modified k-means clustering algorithm to build user profiles as the representatives of usage patterns. Moreover, the hidden task model is derived by characterizing the meaningful latent factor space. With the discovered user profiles, we then choose the most matched profile, which possesses the closely similar preference to current user and make collaborative recommendation based on the corresponding page weights appeared in the selected user profile. The preliminary experimental results performed on real world data sets show that the proposed approach is capable of making recommendation accurately and efficiently.

Discovery driven analysis on semi-structured text data

Discovery Driven Analysis (DDA) is a common feature of OLAP technology to analyze structured data. In essence, DDA helps analysts to discover anomalous data by highlighting 'unexpected' values in the OLAP cube. By giving indications to the analyst on what dimensions to explore, DDA speeds up the process of discovering anomalies and their causes. However, Discovery Driven Analysis (and OLAP in general) is only applicable on structured data, such as records in databases. We propose a system to extend DDA technology to semi-structured text documents, that is, text documents with a few structured data. Our system pipeline consists of two stages: first, the text part of each document is structured around user specified dimensions, using semi-PLSA algorithm; then, we adapt DDA to these fully structured documents, thus enabling DDA on text documents. We present some applications of this system in OLAP analysis and show how scalability issues are solved. Results show that our system can handle reasonable datasets of documents, in real time, without any need for pre-computation.

Influence of Nitrogen and Sink Competition on Shoot Growth of Poplar

Terrestrial and oceanic biomass carbon sinks help reduce anthropogenic CO2 emissions and mitigate the long-term effect of increasing atmospheric CO2. Woody plants have large carbon pools because of their long residence time, however N availability can negatively impact tree responses to elevated CO2. Seasonal cycling of internal N in trees is a component that contributes to fitness especially in N limited environments. It involves resorption from senescing leaves of deciduous trees and storage as vegetative storage proteins (VSP) in perennial organs. Populus is a model organism for tree biology that efficiently recycles N. Bark storage proteins (BSP) are the most abundant VSP that serves as seasonal N reserves. Here I show how poplar growth is influenced by N availability and how growth is influenced by shoot competition for stored N reserves. I also provide data that indicates that auxin mediates BSP catabolism during renewed shoot growth. Understanding the components of N accumulation, remobilization and utilization can provide insights leading to increasing N use efficiency (NUE) of perennial plants.

EFFECT OF COMPOST ON THE MICROCLOVER ESTABLISHMENT AND USE OF COMPOST AND MICROCLOVER TO REDUCE LAWN NUTRIENT RUNOFF

High volume compost incorporation can reduce runoff from compacted soils but its use also associated with elevated N and P concentrations in runoff making it difficult to assess if this practice will reduce nutrient loading of surface waters. Additionally, little is known about how this practice will effect leguminous species establishment in lawns as means to reduce long term fertilizer use. When 5 cm of compost was incorporated into soil a reduction in runoff of 40 and 59% was needed for N and P losses from a tall fescue + microclover lawn to be equivalent to a non-compost amended soil supporting a well fertilized tall fescue lawn. Use of compost as a soil amendment resulted in quicker lawn establishment and darker color, when compared to non-amended soil receiving a mineral fertilizer. Biosolid composts containing high amounts ammonium severely reduce the establishment of clover in tall fescue + micrclover seed mixture.

Regenerative Stormwater Conveyance: Design Implications of an Urban Case Demonstration in Baltimore, Maryland

This research-design thesis explores the implementation of Regenerative Stormwater Conveyance (RSC) as a retrofit of an existing impervious drainage system in a small catchment in the degraded Jones Falls watershed in Baltimore City. An introduction to RSC is provided, placing its development within a theoretical context of novel ecosystems, biomimicry and Nassauer and Opdam’s (2008) model of landscape innovation. The case site is in Baltimore’s Hampden neighborhood on City-owned land adjacent to rowhomes, open space and an access point to a popular wooded trail along a local stream. The design proposal employs RSC to retrofit an ill-performing stormwater system, simultaneously providing a range of ecological, social and economic services; water quantity, water quality and economic performance of the proposed RSC are quantified. While the proposed design is site-specific the model is adaptable for retrofitting other small-scale impervious drainage systems, providing a strategic tool in addressing Baltimore City’s stormwater challenges.

MAXIMIZING LANDSCAPE PERFORMANCE AT ADVENTIST HOSPITAL: HEALING THE PEOPLE, HEALING SLIGO CREEK

This paper answers the question of whether a design intervention on Washington Adventist Hospital’s Takoma Park campus can combine stormwater Best Management Practices with outdoor healing spaces, to improve the health of the local creek (Sligo Creek) while creating a restorative environment for the hospital community. To improve the health of Sligo Creek, a campus-wide stormwater analysis was undertaken, in addition to an intervention-site-specific stormwater analysis, and a literature review of stormwater best management practices. To create a restorative environment, a literature review of healing gardens was undertaken, in addition to a campus-wide site analysis, to uncover the most ideally suited site to create a restorative environment.

SHIFTING SCALES, ADJUSTING LENSES: A FRAMEWORK FOR INVESTIGATING BALTIMORE’S URBAN VACANCY

This thesis addresses contemporary gaps of vacancy within literature by using qualitative and quantitative methods and tools to determine the quantity, location, and interspatial relationships of vacant buildings and lots located in Baltimore Maryland. Spatial analyses were conducted to answer three questions of vacancy: 1) how many vacant lots and buildings exist, 2) whether there are spatial patterns of vacancy, such as clustering around geographic locations or within watersheds, and 3) how to prioritize intervention opportunities that respond to the city's larger issues? Using the city’s vacant lot and building data-sets, two concepts emerged from these investigations. First, Utilized Landscapes as a classification system that identifies lands that serve a function but have un-traditional qualities that make them susceptible to being labeled “vacant.” Second, the development of Transitional Zones, geographical areas with a high density of vacant buildings or lots that should be prioritized.