In this paper a new approach to metadata production is pre- sented. For this purpose, a new inter... more In this paper a new approach to metadata production is pre- sented. For this purpose, a new interactive tool for audio- visual content acquisition and classication has been devel- oped. The user can decompose a given content into units and easily annotate each unit adding basic information such as time, place, etc. as well classication information such as event type, relationship type, etc. according to the MPEGñ 7 Standard. At the end of this production process, the tool automatically produces a structural description of the over- all set of the annotated units. The new idea proposed in this work is to combine the intrinsic semantics of each annotated unit with the implicit semantic information derived from the structural description, hence reducing the needs to perform complex signal processing operations on the content. This aspect is really important since image and video process- ing is generally heavier than metadata processing and the content can be spread over a network and ...
Context and objectives The characteristics of PCs, huge storage capacity, tremendous processing p... more Context and objectives The characteristics of PCs, huge storage capacity, tremendous processing power, and high flexibility are becoming available for consumer devices like set-top boxes, TVs, and VCRs. Interconnection of these devices and wireless communication with various portable devices will create a complex home system with the capacity to store many types of data and offer new ways of interacting with it. To offer the user high flexibility and ease of use, new solutions are required. Advanced retrieval methods are needed to support accessing data stored anywhere in the home system from any device. Meta-data obtained through analysis, services, and logging user behaviour is needed to support these functions. Transfer of data must be easy, and transfer and adaptation of accompanying meta-data must be transparent to the user. The combination of broadcast, storage, and internet will open the way to new types of applications and interactions with the home system. A large distribut...
Recently, a new type of camera, which allows the recording of almost the entire range of luminosi... more Recently, a new type of camera, which allows the recording of almost the entire range of luminosity of the acquired scene, has been introduced to the market. The produced High Dynamic Range images (HDRi), can simplify some video-surveillance tasks, such as object detection, recognition and tracking, since they can simultaneously capture the visual scene content of both dark and bright areas. Applications where lighting conditions cannot be controlled can greatly benefit from the adoption of HDR cameras. For example, the achievement of a good and constant visual quality of images is one of the key aspects in video-surveillance applications, because it allows to robustly de-tect salient events in the captured scenes. However, the use of HDR images in traditional video-surveillance systems re-quires the reduction of their dynamic range appropriately, by applying a tone mapping operator. In this paper a fast method for tone mapping HDR videos, which combines the benefits of both local a...
2000 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings (Cat. No.00CH37100), 2000
Page 1. DESCRIBING MULTIMEDIA DOCUMENTS IN NATURAL AND SEMANTIC-DRIVEN ORDERED HIERARCHIES N. Ada... more Page 1. DESCRIBING MULTIMEDIA DOCUMENTS IN NATURAL AND SEMANTIC-DRIVEN ORDERED HIERARCHIES N. Adanzi, A. Bugntti, R. Leonnrdi, P. Migliornti, Loi-etzzo A. Rossi Department of Electronics for Automation ...
2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2012
ABSTRACT This paper proposes a two-layer High Dynamic Range (HDR) coding scheme using a new tone ... more ABSTRACT This paper proposes a two-layer High Dynamic Range (HDR) coding scheme using a new tone mapping. Our tone mapping method transforms an HDR image onto a Low Dynamic Range (LDR) image by using a base map that is a smoothed version of the HDR luminance. In our scheme, the HDR image can be reconstructed from the tone mapped LDR image. Our method makes use of this property to realize a two-layer HDR coding by encoding both of the tone mapped LDR image and the base map. This paper validates its effectiveness of our approach through some experiments.
As the High dynamic range (HDR) images generally have more than 10 bit/1024 colors per channel, i... more As the High dynamic range (HDR) images generally have more than 10 bit/1024 colors per channel, its enormous data size often needs to be reduced when transmitting or storing the images. Thus development for functional compression is one of important research topics. Recently a lot of techniques for the HDR image compression are being suggested, and several two-layer coding algorithms which separately encode a low dynamic range (LDR) image and a residual image have been studied. However, those methods are inefficient in coding performance. In this article, we suggest a new two-layer coding algorithm for the HDR images, which realizes two-layered dynamic range. Our method encodes a base map, which is a blurred version of the HDR, and LDR image produced by the base map. Our algorithm significantly improves a compression performance.
2011 18th IEEE International Conference on Image Processing, 2011
Content descriptors, useful for browsing and retrieval tasks, are generally extracted and treated... more Content descriptors, useful for browsing and retrieval tasks, are generally extracted and treated as a separate entity with respect to the nature of the content itself. At the same time, conventional coding processes do not take into account information carried out by content descriptors. Content descriptors are closely related to the content itself, and they potentially can be used to
2013 IEEE 15th International Workshop on Multimedia Signal Processing (MMSP), 2013
ABSTRACT In this paper, we propose a methodology to allow movie character recognition and trackin... more ABSTRACT In this paper, we propose a methodology to allow movie character recognition and tracking within movie scenes. In detail, we present a combination of a tracking algorithm robust against the problems of the currently available face detection algorithms and a face recognition process. We test how effective the system is in terms of both face tracking effectiveness and precision-recall results obtained for the recognition of the main characters present in an input movie.
In this paper a new approach to metadata production is pre- sented. For this purpose, a new inter... more In this paper a new approach to metadata production is pre- sented. For this purpose, a new interactive tool for audio- visual content acquisition and classication has been devel- oped. The user can decompose a given content into units and easily annotate each unit adding basic information such as time, place, etc. as well classication information such as event type, relationship type, etc. according to the MPEGñ 7 Standard. At the end of this production process, the tool automatically produces a structural description of the over- all set of the annotated units. The new idea proposed in this work is to combine the intrinsic semantics of each annotated unit with the implicit semantic information derived from the structural description, hence reducing the needs to perform complex signal processing operations on the content. This aspect is really important since image and video process- ing is generally heavier than metadata processing and the content can be spread over a network and ...
Context and objectives The characteristics of PCs, huge storage capacity, tremendous processing p... more Context and objectives The characteristics of PCs, huge storage capacity, tremendous processing power, and high flexibility are becoming available for consumer devices like set-top boxes, TVs, and VCRs. Interconnection of these devices and wireless communication with various portable devices will create a complex home system with the capacity to store many types of data and offer new ways of interacting with it. To offer the user high flexibility and ease of use, new solutions are required. Advanced retrieval methods are needed to support accessing data stored anywhere in the home system from any device. Meta-data obtained through analysis, services, and logging user behaviour is needed to support these functions. Transfer of data must be easy, and transfer and adaptation of accompanying meta-data must be transparent to the user. The combination of broadcast, storage, and internet will open the way to new types of applications and interactions with the home system. A large distribut...
Recently, a new type of camera, which allows the recording of almost the entire range of luminosi... more Recently, a new type of camera, which allows the recording of almost the entire range of luminosity of the acquired scene, has been introduced to the market. The produced High Dynamic Range images (HDRi), can simplify some video-surveillance tasks, such as object detection, recognition and tracking, since they can simultaneously capture the visual scene content of both dark and bright areas. Applications where lighting conditions cannot be controlled can greatly benefit from the adoption of HDR cameras. For example, the achievement of a good and constant visual quality of images is one of the key aspects in video-surveillance applications, because it allows to robustly de-tect salient events in the captured scenes. However, the use of HDR images in traditional video-surveillance systems re-quires the reduction of their dynamic range appropriately, by applying a tone mapping operator. In this paper a fast method for tone mapping HDR videos, which combines the benefits of both local a...
2000 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings (Cat. No.00CH37100), 2000
Page 1. DESCRIBING MULTIMEDIA DOCUMENTS IN NATURAL AND SEMANTIC-DRIVEN ORDERED HIERARCHIES N. Ada... more Page 1. DESCRIBING MULTIMEDIA DOCUMENTS IN NATURAL AND SEMANTIC-DRIVEN ORDERED HIERARCHIES N. Adanzi, A. Bugntti, R. Leonnrdi, P. Migliornti, Loi-etzzo A. Rossi Department of Electronics for Automation ...
2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2012
ABSTRACT This paper proposes a two-layer High Dynamic Range (HDR) coding scheme using a new tone ... more ABSTRACT This paper proposes a two-layer High Dynamic Range (HDR) coding scheme using a new tone mapping. Our tone mapping method transforms an HDR image onto a Low Dynamic Range (LDR) image by using a base map that is a smoothed version of the HDR luminance. In our scheme, the HDR image can be reconstructed from the tone mapped LDR image. Our method makes use of this property to realize a two-layer HDR coding by encoding both of the tone mapped LDR image and the base map. This paper validates its effectiveness of our approach through some experiments.
As the High dynamic range (HDR) images generally have more than 10 bit/1024 colors per channel, i... more As the High dynamic range (HDR) images generally have more than 10 bit/1024 colors per channel, its enormous data size often needs to be reduced when transmitting or storing the images. Thus development for functional compression is one of important research topics. Recently a lot of techniques for the HDR image compression are being suggested, and several two-layer coding algorithms which separately encode a low dynamic range (LDR) image and a residual image have been studied. However, those methods are inefficient in coding performance. In this article, we suggest a new two-layer coding algorithm for the HDR images, which realizes two-layered dynamic range. Our method encodes a base map, which is a blurred version of the HDR, and LDR image produced by the base map. Our algorithm significantly improves a compression performance.
2011 18th IEEE International Conference on Image Processing, 2011
Content descriptors, useful for browsing and retrieval tasks, are generally extracted and treated... more Content descriptors, useful for browsing and retrieval tasks, are generally extracted and treated as a separate entity with respect to the nature of the content itself. At the same time, conventional coding processes do not take into account information carried out by content descriptors. Content descriptors are closely related to the content itself, and they potentially can be used to
2013 IEEE 15th International Workshop on Multimedia Signal Processing (MMSP), 2013
ABSTRACT In this paper, we propose a methodology to allow movie character recognition and trackin... more ABSTRACT In this paper, we propose a methodology to allow movie character recognition and tracking within movie scenes. In detail, we present a combination of a tracking algorithm robust against the problems of the currently available face detection algorithms and a face recognition process. We test how effective the system is in terms of both face tracking effectiveness and precision-recall results obtained for the recognition of the main characters present in an input movie.
Uploads
Papers