Many algorithms have been proposed in literature for digital film restoration; unfortunately, none of them ensures a perfect restoration whichever is the image sequence to be restored. In this paper, we adopt an approach to digital scratch restoration based on the combination of relatively well settled distinct techniques. We propose P-LSR, a parallel version focusing on strategies based on both task and data partitioning to achieve good load balancing, that well adapts also to be distributed. Parallel performance is deeply investigated for several real image sequences.
{"title":"P-LSR: a parallel algorithm for line scratch restoration","authors":"G. Laccetti, L. Maddalena, A. Petrosino","doi":"10.1109/CAMP.2005.37","DOIUrl":"https://doi.org/10.1109/CAMP.2005.37","url":null,"abstract":"Many algorithms have been proposed in literature for digital film restoration; unfortunately, none of them ensures a perfect restoration whichever is the image sequence to be restored. In this paper, we adopt an approach to digital scratch restoration based on the combination of relatively well settled distinct techniques. We propose P-LSR, a parallel version focusing on strategies based on both task and data partitioning to achieve good load balancing, that well adapts also to be distributed. Parallel performance is deeply investigated for several real image sequences.","PeriodicalId":393875,"journal":{"name":"Seventh International Workshop on Computer Architecture for Machine Perception (CAMP'05)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2005-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128683200","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Many image processing algorithms have a very high execution time if only a processor is used for processing them. Using a SIMD parallel structure for its execution could reduce this time. This is particularly important in the case of algorithms that must be processed in real time. The use of networks of PC is an appealing solution that besides its low cost, takes advantage from both the high speed of actual interconnection networks, and the high-performance of PC. In this paper we present a model that explicitly considers system parameters, network parameters, and application parameters. So, the speed and communication model of the considered network, the workstations and PC computing power, the per-pixel computational cost of the algorithms (that can be constant or variable), and a variable number of computers have been considered. We do not aim to evaluate the processing of high and medium-level algorithms of a MISD structure, but we present the first results of our evaluations for iterative low-level image processing applications. Specifically, we give a prediction model to distribute the data to each processor of a distributed system, minimizing the processors' idle time.
{"title":"On real time image processing on a network of PCs","authors":"P. Millán, E. Montseny","doi":"10.1109/CAMP.2005.35","DOIUrl":"https://doi.org/10.1109/CAMP.2005.35","url":null,"abstract":"Many image processing algorithms have a very high execution time if only a processor is used for processing them. Using a SIMD parallel structure for its execution could reduce this time. This is particularly important in the case of algorithms that must be processed in real time. The use of networks of PC is an appealing solution that besides its low cost, takes advantage from both the high speed of actual interconnection networks, and the high-performance of PC. In this paper we present a model that explicitly considers system parameters, network parameters, and application parameters. So, the speed and communication model of the considered network, the workstations and PC computing power, the per-pixel computational cost of the algorithms (that can be constant or variable), and a variable number of computers have been considered. We do not aim to evaluate the processing of high and medium-level algorithms of a MISD structure, but we present the first results of our evaluations for iterative low-level image processing applications. Specifically, we give a prediction model to distribute the data to each processor of a distributed system, minimizing the processors' idle time.","PeriodicalId":393875,"journal":{"name":"Seventh International Workshop on Computer Architecture for Machine Perception (CAMP'05)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2005-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122313988","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper describes a task-independent controller that allows for an easy implementation of vision systems for processing video sequences. The controller does not have a fixed dataflow or any fixed steps. The dataflow is constructed by the modules by describing themselves for the controller. During operation the modules and their parameters are selected using an independent decision module. This makes the system flexible and allows comparison of different learning techniques and decision strategies. The controller is being used by the CAVIAR system and its current decision module is a rule-based system written in Clips.
{"title":"An intelligent and task-independent controller for video sequence analysis","authors":"J. Bins, T. List, Robert B. Fisher, David Tweed","doi":"10.1109/CAMP.2005.12","DOIUrl":"https://doi.org/10.1109/CAMP.2005.12","url":null,"abstract":"This paper describes a task-independent controller that allows for an easy implementation of vision systems for processing video sequences. The controller does not have a fixed dataflow or any fixed steps. The dataflow is constructed by the modules by describing themselves for the controller. During operation the modules and their parameters are selected using an independent decision module. This makes the system flexible and allows comparison of different learning techniques and decision strategies. The controller is being used by the CAVIAR system and its current decision module is a rule-based system written in Clips.","PeriodicalId":393875,"journal":{"name":"Seventh International Workshop on Computer Architecture for Machine Perception (CAMP'05)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2005-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130965200","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We have simulated the implementation of 16-bit floating point instructions on a Pentium4 and PowerPC G4 and G5 to evaluate the performance impact of these instructions in embedded processors for graphics and multimedia applications. Both accuracy of the computations and the execution time have been considered. For low-end embedded processors, the 16-bit FP instructions deliver a larger dynamic range than 16-bit integer with the same memory footprint. For high-end embedded processors, we add the speed up coming from wider SIMD instructions.
{"title":"16-bit floating point instructions for embedded multimedia applications","authors":"L. Lacassagne, D. Etiemble, S. A. O. Kablia","doi":"10.1109/CAMP.2005.1","DOIUrl":"https://doi.org/10.1109/CAMP.2005.1","url":null,"abstract":"We have simulated the implementation of 16-bit floating point instructions on a Pentium4 and PowerPC G4 and G5 to evaluate the performance impact of these instructions in embedded processors for graphics and multimedia applications. Both accuracy of the computations and the execution time have been considered. For low-end embedded processors, the 16-bit FP instructions deliver a larger dynamic range than 16-bit integer with the same memory footprint. For high-end embedded processors, we add the speed up coming from wider SIMD instructions.","PeriodicalId":393875,"journal":{"name":"Seventh International Workshop on Computer Architecture for Machine Perception (CAMP'05)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2005-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132309086","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper presents an architecture for cognitive analysis of streaming video, in which a new module can easily be plugged in, to add to or even compete with existing functionality. This allows the implementers to focus on the key scientific issues instead of struggling with the details of the implementation. The architecture is distributed and runs independently of the underlying computer architecture and can run transparently across one or many different operating systems in a larger distributed system. This architecture focuses on several key computer vision issues, such as multi-level global and local control, automatic dataflow based on auto-descriptive self-regulating independent modules that come together to form a whole based on the characteristics of the individual and the needs of the system rather than a static flow diagram.
{"title":"A plug-and-play architecture for cognitive video stream analysis","authors":"T. List, J. Bins, Robert B. Fisher, David Tweed","doi":"10.1109/CAMP.2005.5","DOIUrl":"https://doi.org/10.1109/CAMP.2005.5","url":null,"abstract":"This paper presents an architecture for cognitive analysis of streaming video, in which a new module can easily be plugged in, to add to or even compete with existing functionality. This allows the implementers to focus on the key scientific issues instead of struggling with the details of the implementation. The architecture is distributed and runs independently of the underlying computer architecture and can run transparently across one or many different operating systems in a larger distributed system. This architecture focuses on several key computer vision issues, such as multi-level global and local control, automatic dataflow based on auto-descriptive self-regulating independent modules that come together to form a whole based on the characteristics of the individual and the needs of the system rather than a static flow diagram.","PeriodicalId":393875,"journal":{"name":"Seventh International Workshop on Computer Architecture for Machine Perception (CAMP'05)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2005-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123580395","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper we describe a two step algorithm which localizes faces in 2D color images depicting a single face on a complex background. Given a single image, the algorithm roughly determines the skin regions and then searches for eyes within them. A face is localized if at least one eye is present in a skin region. The system is based on a support vector machine trained to separate sub-images representing eyes from others. The algorithm is robust to scale, illumination, pose variations and deals with partial occlusions. Results on several public databases are presented.
{"title":"Face localization in color images with complex background","authors":"P. Campadelli, R. Lanzarotti, G. Lipori","doi":"10.1109/CAMP.2005.27","DOIUrl":"https://doi.org/10.1109/CAMP.2005.27","url":null,"abstract":"In this paper we describe a two step algorithm which localizes faces in 2D color images depicting a single face on a complex background. Given a single image, the algorithm roughly determines the skin regions and then searches for eyes within them. A face is localized if at least one eye is present in a skin region. The system is based on a support vector machine trained to separate sub-images representing eyes from others. The algorithm is robust to scale, illumination, pose variations and deals with partial occlusions. Results on several public databases are presented.","PeriodicalId":393875,"journal":{"name":"Seventh International Workshop on Computer Architecture for Machine Perception (CAMP'05)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2005-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131602272","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We present a Web-based system for remotely reconstructing 3D models from 2D images. The system relies on a full static stereo vision procedure, which includes feature extraction; matching; fundamental matrix and roto-translation parameters estimation; image rectification; 3D-point triangulation and poly-hedrization, to produce a 3D geometrical model approximating the input scene. The system architecture, grounded on a conventional client-server scheme, has been designed according to the model-view-controller (MVC) paradigm in order to obtain a modular configuration, keeping separate the presentation from the processing and control phases. The resulting, flexible scheme allows the user to fully control the flow of information by means of an interactive interface. In this way, the partial results arising from the numerical steps can be refined by the user, who can modify input parameters, add/eliminate/change output results, thereby actively contributing to an accurate and efficient synthesis of the final model. The system has been realized on an open-source platform; results are presented and discussed.
{"title":"3D reconstruction as a network service","authors":"S. Ansoldi, Giovanni Battista Buora, V. Roberto","doi":"10.1109/CAMP.2005.2","DOIUrl":"https://doi.org/10.1109/CAMP.2005.2","url":null,"abstract":"We present a Web-based system for remotely reconstructing 3D models from 2D images. The system relies on a full static stereo vision procedure, which includes feature extraction; matching; fundamental matrix and roto-translation parameters estimation; image rectification; 3D-point triangulation and poly-hedrization, to produce a 3D geometrical model approximating the input scene. The system architecture, grounded on a conventional client-server scheme, has been designed according to the model-view-controller (MVC) paradigm in order to obtain a modular configuration, keeping separate the presentation from the processing and control phases. The resulting, flexible scheme allows the user to fully control the flow of information by means of an interactive interface. In this way, the partial results arising from the numerical steps can be refined by the user, who can modify input parameters, add/eliminate/change output results, thereby actively contributing to an accurate and efficient synthesis of the final model. The system has been realized on an open-source platform; results are presented and discussed.","PeriodicalId":393875,"journal":{"name":"Seventh International Workshop on Computer Architecture for Machine Perception (CAMP'05)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2005-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128092383","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper we address the problem of architectural exploration from the energy/performance point of view of a VLIW processor for embedded systems. We also consider an architectural modification we introduced in order to extend the reference processor so that it can exploit both instruction level parallelism and thread level parallelism. A power model obtained by applying an instruction-level power estimation technique is presented and validated with experimental results. This power model was plugged in a parametric cycle-accurate simulator in order to support architectural exploration. Experimental results derived from the proposed framework show a comparison among different implementations of the reference processor: single and dual cluster implementations, and dual cluster with multithreaded extension.
{"title":"Energy/performance evaluation of the multithreaded extension of a multicluster VLIW processor","authors":"D. Barretta, G. Palermo, M. Sami, R. Zafalon","doi":"10.1109/CAMP.2005.25","DOIUrl":"https://doi.org/10.1109/CAMP.2005.25","url":null,"abstract":"In this paper we address the problem of architectural exploration from the energy/performance point of view of a VLIW processor for embedded systems. We also consider an architectural modification we introduced in order to extend the reference processor so that it can exploit both instruction level parallelism and thread level parallelism. A power model obtained by applying an instruction-level power estimation technique is presented and validated with experimental results. This power model was plugged in a parametric cycle-accurate simulator in order to support architectural exploration. Experimental results derived from the proposed framework show a comparison among different implementations of the reference processor: single and dual cluster implementations, and dual cluster with multithreaded extension.","PeriodicalId":393875,"journal":{"name":"Seventh International Workshop on Computer Architecture for Machine Perception (CAMP'05)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2005-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129144267","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
All sciences, including astronomy, are now entering the era of information abundance. The exponentially increasing volume and complexity of modern data sets promises to transform the scientific practice, but also poses a number of common technological challenges. The virtual observatory concept is the astronomical community's response to these challenges: it aims to harness the progress in information technology in the service of astronomy, and at the same time provide a valuable testbed for information technology and applied computer science. Challenges broadly fall into two categories: data handling (or "data farming"), including issues such as archives, intelligent storage, databases, interoperability, fast networks, etc., and data mining, data understanding, and knowledge discovery, which include issues such as automated clustering and classification, multivariate correlation searches, pattern recognition, visualization in highly hyperdimensional parameter spaces, etc., as well as various applications of machine learning in these contexts. Such techniques are forming a methodological foundation for science with massive and complex data sets in general, and are likely to have a much broader impact on the modern society, commerce, information economy, security, etc. There is a powerful emerging synergy between the computationally enabled science and the science-driven computing, which will drive the progress in science, scholarship, and many other venues in the 21st century.
{"title":"Virtual astronomy, information technology, and the new scientific methodology","authors":"S. Djorgovski","doi":"10.1109/CAMP.2005.53","DOIUrl":"https://doi.org/10.1109/CAMP.2005.53","url":null,"abstract":"All sciences, including astronomy, are now entering the era of information abundance. The exponentially increasing volume and complexity of modern data sets promises to transform the scientific practice, but also poses a number of common technological challenges. The virtual observatory concept is the astronomical community's response to these challenges: it aims to harness the progress in information technology in the service of astronomy, and at the same time provide a valuable testbed for information technology and applied computer science. Challenges broadly fall into two categories: data handling (or \"data farming\"), including issues such as archives, intelligent storage, databases, interoperability, fast networks, etc., and data mining, data understanding, and knowledge discovery, which include issues such as automated clustering and classification, multivariate correlation searches, pattern recognition, visualization in highly hyperdimensional parameter spaces, etc., as well as various applications of machine learning in these contexts. Such techniques are forming a methodological foundation for science with massive and complex data sets in general, and are likely to have a much broader impact on the modern society, commerce, information economy, security, etc. There is a powerful emerging synergy between the computationally enabled science and the science-driven computing, which will drive the progress in science, scholarship, and many other venues in the 21st century.","PeriodicalId":393875,"journal":{"name":"Seventh International Workshop on Computer Architecture for Machine Perception (CAMP'05)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2005-04-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122869360","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}