|
|
Similarity search in image collections
Navrátil, Lukáš ; Bartoš, Tomáš (advisor) ; Skopal, Tomáš (referee)
Detection of keypoints from image and their characterization by using descriptors is common technique in some branches of computer vision. The goal of this thesis is to explore and confirm usability of this technique for similarity retrieval in image collections. For this purpose it will be created a web application used for collecting ratings of similarity from users which will be subsequently compared with results computed by the implementation of SURF algorithm, one of algorithms used for detection and description of image keypoints. It will also be discussed the impact of metrics and parameters influencing results of computation of similarity between images and it will be made an effort to find settings for which computed results will be closest to user's similarity perception.
|
|
|
Similarity search in Mass Spectra Databases
Novák, Jiří ; Skopal, Tomáš (advisor) ; Svozil, Daniel (referee) ; Nahnsen, Sven (referee)
Shotgun proteomics is a widely known technique for identification of protein and peptide sequences from an "in vitro" sample. A tandem mass spectrometer generates tens of thousands of mass spectra which must be annotated with peptide sequences. For this purpose, the similarity search in a database of theoretical spectra generated from a database of known protein sequences can be utilized. Since the sizes of databases grow rapidly in recent years, there is a demand for utilization of various database indexing techniques. We investigate the capabilities of (non)metric access methods as the database indexing techniques for fast and approximate similarity retrieval in mass spectra databases. We show that the method for peptide sequences identification is more than 100x faster than a sequential scan over the entire database while more than 90% of spectra are correctly annotated with peptide sequences. Since the method is currently suitable for small mixtures of proteins, we also utilize a precursor mass filter as the database indexing technique for complex mixtures of proteins. The precursor mass filter followed by ranking of spectra by a modification of the parametrized Hausdorff distance outperforms state-of-the-art tools in the number of identified peptide sequences and the speed of search. The...
|
|
|
Similarity Search in Protein Structure Databases
Galgonek, Jakub ; Skopal, Tomáš (advisor) ; Porto, Markus (referee) ; Svozil, Daniel (referee)
Proteins are one of the most important biopolymers having a wide range of functions in living organisms. Their huge functional diversity is achieved by their ability to fold into various 3D structures. Moreover, it has been shown that proteins sharing similar structure often share also other properties (e.g, a biological function, an evolutionary origin, etc.). Therefore, protein structures and methods to identify their similarities are so widely studied. In this thesis, we introduce a system allowing similarity search in pro- tein structure databases. The system retrieves, given a query structure, all database structures being similar to the query structure. It employs several key components. We have introduced a novel similarity measure assigning similarity scores to pairs of protein structures. We have designed specific access method based on LAESA metric indexing and using the proposed measure. The access method allows to search similar structures more effi- ciently than when a sequential scan of a database is employed. To achieve further speedup, the measure and the access method have been parallelized, resulting in almost linear speedup with the respect to the number of available cores. The last component is a web user interface that allows to accept a query structure and to present a list of...
|
|
|
Employing Parallel Architectures in Similarity Search
Kruliš, Martin ; Yaghob, Jakub (advisor) ; Platoš, Jan (referee) ; Pllana, Sabri (referee)
This work examines the possibilities of employing highly parallel architectures in database systems, which are based on the similarity search paradigm. The main objective of our research is utilizing the computational power of current GPU devices for similarity search in the databases of images. Despite leaping progress made in the past few years, the similarity search problems remain very expensive from a compu- tational point of view, which limits the scope of their applicability. GPU devices have a tremendous computational power at their disposal; however, the usability of this power for particular problems is often complicated due to the specific properties of this architecture. Therefore, the existing algorithms and data structures require extensive modifications if they are to be adapted for the GPUs. We have addressed all the aspects of this domain, such as efficient utilization of the GPU hardware for generic computations, parallelization of similarity search process, and acceleration of image indexing techniques. In most cases, employing the GPU devices brought a speedup of two orders of magnitude with respect to single-core CPUs and approximately one order of magnitude with respect to multiprocessor NUMA servers. This thesis summarizes our experience and discoveries from several years of research,...
|
|
|
Similarity Search in Protein Structure Databases
Galgonek, Jakub
Proteins are one of the most important biopolymers having a wide range of functions in living organisms. Their huge functional diversity is achieved by their ability to fold into various 3D structures. Moreover, it has been shown that proteins sharing similar structure often share also other properties (e.g, a biological function, an evolutionary origin, etc.). Therefore, protein structures and methods to identify their similarities are so widely studied. In this thesis, we introduce a system allowing similarity search in pro- tein structure databases. The system retrieves, given a query structure, all database structures being similar to the query structure. It employs several key components. We have introduced a novel similarity measure assigning similarity scores to pairs of protein structures. We have designed specific access method based on LAESA metric indexing and using the proposed measure. The access method allows to search similar structures more effi- ciently than when a sequential scan of a database is employed. To achieve further speedup, the measure and the access method have been parallelized, resulting in almost linear speedup with the respect to the number of available cores. The last component is a web user interface that allows to accept a query structure and to present a list of...
|
|
|
Content-based exploration of unstructured data
Čech, Přemysl ; Lokoč, Jakub (advisor) ; Barthel, Kai Uwe (referee) ; Gudmundsson, Gylfi Thor (referee)
Effective analysis, searching and browsing throughout arbitrary multimedia collections is still a challenging task. To perform a search among multimedia objects, first, a similarity model has to be defined. Such a model establishes methods describing how the content of individual objects is processed and how key features and descriptors, that are used for modeling similarity between objects, are formed. This task is not trivial since there can be many ways of determining how to comprehend the content of multimedia data. Furthermore, with the growing size of contemporary database collections, multimedia retrieval and exploration are extremely computationally intensive. Hence, researchers investigate support indexing structures that can evaluate similarity queries and can respond to user's queries in almost real-time even on datasets counting billions of objects. Another very important aspect of a retrieval system is the user interface for defining queries as well as presenting retrieved results. A multimedia system should offer various inputs for formulating user's queries, especially for situations in which a user cannot provide an ideal query example. Finally, a well- arranged and easy to read interface for visualization of retrieved results is essential for the success of a multimedia exploration and...
|
|
|
Comparison of signature-based and semantic similarity models
Kovalčík, Gregor ; Lokoč, Jakub (advisor) ; Mráz, František (referee)
Content-based image retrieval and similarity search has been investigated for several decades with many different approaches proposed. This thesis fo- cuses on a comparison of two orthogonal similarity models on two different im- age retrieval tasks. More specifically, traditional image representation models based on feature signatures are compared with models based on state-of-the-art deep convolutional neural networks. Query-by-example benchmarking and tar- get browsing tasks were selected for the comparison. In a thorough experimental evaluation, we confirm that models based on deep convolutional neural networks outperform the traditional models. However, in the target browsing scenario, we show that the traditional models could still represent an effective option. We have also implemented a feature signature extractor into the OpenCV library in order to make the source codes available for the image retrieval and computer vision community. 1
|
|
|
The Impact of Image Resolution on the Precision of Content-based Retrieval
Navrátil, Lukáš ; Lokoč, Jakub (advisor) ; Skopal, Tomáš (referee)
This thesis is focused on comparing methods for similarity image retrieval. Common techniques and testing sets are introduced. The testing sets are there to measure the accuracy of the searching systems based on similarity image retrieval. Measurements are done on those models which are implemented on the basis of presented techniques. These measurements examine their results depending on the input data, used components and parameters settings, especially the impact of image resolution on the retrieval precision is examined. These results are analysed and the models are compared. Powered by TCPDF (www.tcpdf.org)
|
|
|
Satellite data management and analysis
Krška, David ; Lokoč, Jakub (advisor) ; Nečaský, Martin (referee)
The aim of this bachelor thesis is to implement and test an interactive application for the analysis, visualization and classification of satellite data. The satellite data are preprocessed and automatically composed into colored images. The application allows to create and compare two types of classification algorithms. The first type uses single images and the second type uses multiple images of the same place, but at different times. We also created an interface for selecting and managing the training data. A few well-known classification algorithms of both types were implemented and their success rates were compared in an experiment. All the satellite data used in the experiment are from the Landsat program. The result of this bachelor thesis is an application primarily focused on classification. But the application could also be extended into a complex GIS system. Powered by TCPDF (www.tcpdf.org)
|
|
|
Employing Parallel Architectures in Similarity Search
Kruliš, Martin ; Yaghob, Jakub (advisor) ; Platoš, Jan (referee) ; Pllana, Sabri (referee)
This work examines the possibilities of employing highly parallel architectures in database systems, which are based on the similarity search paradigm. The main objective of our research is utilizing the computational power of current GPU devices for similarity search in the databases of images. Despite leaping progress made in the past few years, the similarity search problems remain very expensive from a compu- tational point of view, which limits the scope of their applicability. GPU devices have a tremendous computational power at their disposal; however, the usability of this power for particular problems is often complicated due to the specific properties of this architecture. Therefore, the existing algorithms and data structures require extensive modifications if they are to be adapted for the GPUs. We have addressed all the aspects of this domain, such as efficient utilization of the GPU hardware for generic computations, parallelization of similarity search process, and acceleration of image indexing techniques. In most cases, employing the GPU devices brought a speedup of two orders of magnitude with respect to single-core CPUs and approximately one order of magnitude with respect to multiprocessor NUMA servers. This thesis summarizes our experience and discoveries from several years of research,...
|
guest ::
