7th International Conference on Image Processing and its Applications, 1999
It is known that the majority of crime is committed by repeat offenders [3]. It is common for bur... more It is known that the majority of crime is committed by repeat offenders [3]. It is common for burglars to commit a number of offences in the same day. As it would be unusual for an offender to discard their footwear between committing different crimes [4] timely identification and matching of ...
Methodological Issues in Experimental Research in Audiovisual Translation and Media Accessibility
Museums face a particular challenge in enabling blind and partially sighted (BPS) visitors to eng... more Museums face a particular challenge in enabling blind and partially sighted (BPS) visitors to engage emotionally with the narrative(s) they present. In collaboration with a world-leading tourist attraction (Titanic Belfast) and the Royal National Institute of Blind People (RNIB), we have applied several different approaches for improving emotional engagement for BPS visitors. This paper addresses the critical challenge of how to obtain reliable evidence for evaluating the emotional response of BPS visitors to the museum’s audio description (AD) and overall experience. We consider six different methods for measuring emotional engagement, and consider their potential for providing reliable experimental evidence. Based on BPS-user feedback, we present a qualitative comparison of these methods, uniquely all applied to the same museum context.
The success of content-based image finding and retrieval is most marked when the user's requi... more The success of content-based image finding and retrieval is most marked when the user's requirements are very specific. An example of a specific application domain is the grading of engineering materials. In this paper we describe such an application in the area of civil engineering construction materials. We describe an innovative solution to automated vision-based grading of construction materials. From the methodology viewpoint, we show the advantages of using a resolution scale based approach for content characterization of mixtures of fine granularity material and large granularity "aggregate". In particular we use: (i) multiscale entropy; and (ii) significant wavelet coefficients. Links with recent vision model perspectives are discussed.
Many approaches to unconstrained face identification exploit small patches which are unaffected b... more Many approaches to unconstrained face identification exploit small patches which are unaffected by distortions outside of their locality. However, small patches have limited discriminative ability, making accurate patch matching difficult. We propose a novel blockbased approach to exploit the greater discriminative information in larger areas, while maintaining robustness to local variations. A testing block contains several neighbouring testing patches. We identify all the matching training patches in a block jointly, using normalized cross correlation (NCC), as a means of reducing the uncertainty of each matching patch with the addition of the neighbouring patch information. We further propose a multi-scale extension in which we carry out block-based matching at several block sizes, where a larger block contains more neighbouring testing patches, to combine complementary information across scales for further robustness. For evaluation, we use two unconstrained datasets, cropped La...
Discrete wavelet transforms (DWTs) have become a very powerful tool in signal processing. The fir... more Discrete wavelet transforms (DWTs) have become a very powerful tool in signal processing. The first point to be considered if implemented in hardware is what wordlength to use. We detail a mathematical approach that enables the determination of the exact dynamic range at any filter output of a J-stage 1-D and 2-D forward or inverse DWT for any arbitrary input signal. Once applied on the forward transform, our approach leads to wordlength bit savings compared to the work published in the literature. These savings are relevant when implementing the transform in hardware since they lead to considerable area savings and speed improvement. In addition, we provide the first investigation in determining the exact dynamic range of the inverse transform stages. Although our focus is on the DWTs, the approach can be applied on any finite impulse response (FIR)-filter-based architecture.
7th International Conference on Image Processing and its Applications, 1999
It is known that the majority of crime is committed by repeat offenders [3]. It is common for bur... more It is known that the majority of crime is committed by repeat offenders [3]. It is common for burglars to commit a number of offences in the same day. As it would be unusual for an offender to discard their footwear between committing different crimes [4] timely identification and matching of ...
Methodological Issues in Experimental Research in Audiovisual Translation and Media Accessibility
Museums face a particular challenge in enabling blind and partially sighted (BPS) visitors to eng... more Museums face a particular challenge in enabling blind and partially sighted (BPS) visitors to engage emotionally with the narrative(s) they present. In collaboration with a world-leading tourist attraction (Titanic Belfast) and the Royal National Institute of Blind People (RNIB), we have applied several different approaches for improving emotional engagement for BPS visitors. This paper addresses the critical challenge of how to obtain reliable evidence for evaluating the emotional response of BPS visitors to the museum’s audio description (AD) and overall experience. We consider six different methods for measuring emotional engagement, and consider their potential for providing reliable experimental evidence. Based on BPS-user feedback, we present a qualitative comparison of these methods, uniquely all applied to the same museum context.
The success of content-based image finding and retrieval is most marked when the user's requi... more The success of content-based image finding and retrieval is most marked when the user's requirements are very specific. An example of a specific application domain is the grading of engineering materials. In this paper we describe such an application in the area of civil engineering construction materials. We describe an innovative solution to automated vision-based grading of construction materials. From the methodology viewpoint, we show the advantages of using a resolution scale based approach for content characterization of mixtures of fine granularity material and large granularity "aggregate". In particular we use: (i) multiscale entropy; and (ii) significant wavelet coefficients. Links with recent vision model perspectives are discussed.
Many approaches to unconstrained face identification exploit small patches which are unaffected b... more Many approaches to unconstrained face identification exploit small patches which are unaffected by distortions outside of their locality. However, small patches have limited discriminative ability, making accurate patch matching difficult. We propose a novel blockbased approach to exploit the greater discriminative information in larger areas, while maintaining robustness to local variations. A testing block contains several neighbouring testing patches. We identify all the matching training patches in a block jointly, using normalized cross correlation (NCC), as a means of reducing the uncertainty of each matching patch with the addition of the neighbouring patch information. We further propose a multi-scale extension in which we carry out block-based matching at several block sizes, where a larger block contains more neighbouring testing patches, to combine complementary information across scales for further robustness. For evaluation, we use two unconstrained datasets, cropped La...
Discrete wavelet transforms (DWTs) have become a very powerful tool in signal processing. The fir... more Discrete wavelet transforms (DWTs) have become a very powerful tool in signal processing. The first point to be considered if implemented in hardware is what wordlength to use. We detail a mathematical approach that enables the determination of the exact dynamic range at any filter output of a J-stage 1-D and 2-D forward or inverse DWT for any arbitrary input signal. Once applied on the forward transform, our approach leads to wordlength bit savings compared to the work published in the literature. These savings are relevant when implementing the transform in hardware since they lead to considerable area savings and speed improvement. In addition, we provide the first investigation in determining the exact dynamic range of the inverse transform stages. Although our focus is on the DWTs, the approach can be applied on any finite impulse response (FIR)-filter-based architecture.
Uploads
Papers by Danny Crookes