1 resultado para solar to power efficiency
em Boston University Digital Common
Filtro por publicador
- Abertay Research Collections - Abertay University’s repository (3)
- Academic Research Repository at Institute of Developing Economies (2)
- Acceda, el repositorio institucional de la Universidad de Las Palmas de Gran Canaria. España (1)
- AMS Tesi di Dottorato - Alm@DL - Università di Bologna (8)
- AMS Tesi di Laurea - Alm@DL - Università di Bologna (2)
- Andina Digital - Repositorio UASB-Digital - Universidade Andina Simón Bolívar (3)
- Aquatic Commons (5)
- ArchiMeD - Elektronische Publikationen der Universität Mainz - Alemanha (2)
- Archive of European Integration (5)
- Archivo Digital para la Docencia y la Investigación - Repositorio Institucional de la Universidad del País Vasco (1)
- Aston University Research Archive (28)
- Biblioteca Digital da Produção Intelectual da Universidade de São Paulo (3)
- Biblioteca Digital da Produção Intelectual da Universidade de São Paulo (BDPI/USP) (2)
- Biblioteca Digital de la Universidad Católica Argentina (3)
- Biblioteca Digital de Teses e Dissertações Eletrônicas da UERJ (19)
- BORIS: Bern Open Repository and Information System - Berna - Suiça (10)
- Boston University Digital Common (1)
- Brock University, Canada (7)
- Bucknell University Digital Commons - Pensilvania - USA (2)
- CaltechTHESIS (8)
- Cambridge University Engineering Department Publications Database (19)
- CentAUR: Central Archive University of Reading - UK (21)
- Center for Jewish History Digital Collections (8)
- Central European University - Research Support Scheme (1)
- Chinese Academy of Sciences Institutional Repositories Grid Portal (63)
- Cochin University of Science & Technology (CUSAT), India (6)
- Collection Of Biostatistics Research Archive (1)
- Comissão Econômica para a América Latina e o Caribe (CEPAL) (10)
- CORA - Cork Open Research Archive - University College Cork - Ireland (10)
- Dalarna University College Electronic Archive (8)
- Deakin Research Online - Australia (72)
- DI-fusion - The institutional repository of Université Libre de Bruxelles (1)
- Digital Commons - Michigan Tech (6)
- Digital Commons @ DU | University of Denver Research (2)
- Digital Commons at Florida International University (1)
- DigitalCommons@The Texas Medical Center (2)
- Doria (National Library of Finland DSpace Services) - National Library of Finland, Finland (2)
- DRUM (Digital Repository at the University of Maryland) (5)
- Duke University (5)
- Ecology and Society (1)
- eResearch Archive - Queensland Department of Agriculture; Fisheries and Forestry (5)
- Helda - Digital Repository of University of Helsinki (12)
- Indian Institute of Science - Bangalore - Índia (43)
- Instituto Politécnico do Porto, Portugal (20)
- Instituto Superior de Psicologia Aplicada - Lisboa (1)
- Laboratório Nacional de Energia e Geologia - Portugal (1)
- Lume - Repositório Digital da Universidade Federal do Rio Grande do Sul (1)
- Massachusetts Institute of Technology (1)
- National Center for Biotechnology Information - NCBI (4)
- Portal de Revistas Científicas Complutenses - Espanha (1)
- Publishing Network for Geoscientific & Environmental Data (1)
- QSpace: Queen's University - Canada (1)
- QUB Research Portal - Research Directory and Institutional Repository for Queen's University Belfast (52)
- Queensland University of Technology - ePrints Archive (108)
- RDBU - Repositório Digital da Biblioteca da Unisinos (5)
- ReCiL - Repositório Científico Lusófona - Grupo Lusófona, Portugal (1)
- Repositório Aberto da Universidade Aberta de Portugal (1)
- Repositório Científico do Instituto Politécnico de Lisboa - Portugal (3)
- Repositorio de la Universidad de Cuenca (1)
- Repositório digital da Fundação Getúlio Vargas - FGV (33)
- Repositório Institucional da Universidade de Aveiro - Portugal (6)
- Repositório Institucional da Universidade de Brasília (1)
- Repositório Institucional UNESP - Universidade Estadual Paulista "Julio de Mesquita Filho" (124)
- Repositorio Institucional Universidad EAFIT - Medelin - Colombia (1)
- Research Open Access Repository of the University of East London. (3)
- RUN (Repositório da Universidade Nova de Lisboa) - FCT (Faculdade de Cienecias e Technologia), Universidade Nova de Lisboa (UNL), Portugal (5)
- Universidad de Alicante (2)
- Universidad del Rosario, Colombia (18)
- Universidad Politécnica de Madrid (28)
- Universidade de Lisboa - Repositório Aberto (2)
- Universidade Federal do Pará (9)
- Universidade Federal do Rio Grande do Norte (UFRN) (17)
- Universita di Parma (1)
- Universitat de Girona, Spain (1)
- Universitätsbibliothek Kassel, Universität Kassel, Germany (5)
- Université de Lausanne, Switzerland (1)
- Université de Montréal, Canada (18)
- Université Laval Mémoires et thèses électroniques (1)
- University of Canberra Research Repository - Australia (1)
- University of Connecticut - USA (1)
- University of Michigan (6)
- University of Queensland eSpace - Australia (6)
- University of Washington (4)
- WestminsterResearch - UK (4)
- Worcester Research and Publications - Worcester Research and Publications - UK (1)
Resumo:
Many real world image analysis problems, such as face recognition and hand pose estimation, involve recognizing a large number of classes of objects or shapes. Large margin methods, such as AdaBoost and Support Vector Machines (SVMs), often provide competitive accuracy rates, but at the cost of evaluating a large number of binary classifiers, thus making it difficult to apply such methods when thousands or millions of classes need to be recognized. This thesis proposes a filter-and-refine framework, whereby, given a test pattern, a small number of candidate classes can be identified efficiently at the filter step, and computationally expensive large margin classifiers are used to evaluate these candidates at the refine step. Two different filtering methods are proposed, ClassMap and OVA-VS (One-vs.-All classification using Vector Search). ClassMap is an embedding-based method, works for both boosted classifiers and SVMs, and tends to map the patterns and their associated classes close to each other in a vector space. OVA-VS maps OVA classifiers and test patterns to vectors based on the weights and outputs of weak classifiers of the boosting scheme. At runtime, finding the strongest-responding OVA classifier becomes a classical vector search problem, where well-known methods can be used to gain efficiency. In our experiments, the proposed methods achieve significant speed-ups, in some cases up to two orders of magnitude, compared to exhaustive evaluation of all OVA classifiers. This was achieved in hand pose recognition and face recognition systems where the number of classes ranges from 535 to 48,600.