We propose a shape analysis system based upon the description of landmarks with measurement covariance, which extends statistical linear modelling processes to 'pseudo landmarks' for scientific studies.We discuss the properties of our approach and how measurement covariances can be considered characteristic of the local shape.Our formulation includes corrections for parameter bias, induced by the degrees of freedom within the linear model.The method has been implemented and tested on measurements from fly wing, hand and face data.We use these data to explore possible advantages and disadvantages over the use of standard Procrustes/PCA analysis.In the process we show how appropriate weighting provides more efficient use of the available landmark data 1 .

This paper describes a body of multicamera human action video data with manually annotated silhouette data that has been generated for the purpose of evaluating silhouette-based human action recognition methods. It provides a realistic challenge to both the segmentation and human action recognition communities and can act as a benchmark to objectively compare proposed algorithms. The public multi-camera, multi-action dataset is an improvement over existing datasets (e.g. PETS, CAVIAR, soccerdataset) that have not been developed specifically for human action recognition and complements other action recognition datasets (KTH, Weizmann, IXMAS, HumanEva, CMU Motion). It consists of 17 action classes, 14 actors and 8 cameras. Each actor performs an action several times in the action zone. The paper describes the dataset and illustrates a possible approach to algorithm evaluation using a previously published action simple recognition method. In addition to showing an evaluation methodology, these results establish a baseline for other researchers to improve upon.

This paper addresses the problem of silhouette-based human action modeling and recognition, specially when the number of action samples is scarce. The first step of the proposed system is the 2D modeling of human actions based on motion templates, by means of motion history images (MHI). These templates are projected into a new subspace using the Kohonen self organizing feature map (SOM), which groups viewpoint (spatial) and movement (temporal) in a principal manifold, and models the high dimensional space of static templates.The next step is based on the hidden Markov models (HMM) in order to track the map behavior on the temporal sequences of MHI. Every new MHI pattern is compared with the features map obtained during the training. The index of the winner neuron is considered as discrete observation for the HMM. If the number of samples is not enough, a sampling technique, the sampling importance resampling (SIR) algorithm, is applied in order to increase the number of observations for the HMM. Finally, temporal pattern recognition is accomplished by a maximum likelihood (ML) classifier. We demonstrate this approach on two publicly available dataset: one based on real actors and another one based on virtual actors.

We propose a new method for human action recognition from video streams that is fast and robust to noise and to large changes in camera views. We start by extracting features in the Fourier domain once we obtain the bounding boxes containing the silhouettes of a human for a number of video frames representing a basic action. After preprocessing, we divide each space-time volume into space-time sub-volumes (STSV) and compute their corresponding mean-power spectra as our feature vectors. Our features result in high classification performance even with simple distance measures. We perform an experimental comparison, using the same data, between our method and two state-of-the-art methods.

This paper addresses the problem of determining skin reflectance parameters, and studies their stability and discriminating power for different individuals. Our study uses radiance data captured by a Cyberware 3030 range scanner. We analyse the data using a layered reflectance model based on the Beckmann-Kirchhoff wave scattering model. The parameters of this model are the thickness of the skin layers, and their roughness. We investigate how the parameters of this model vary between different subjects, and how they vary for the same subject under different conditions.

Recent work has shown how the Fresnel coefficient can be used to improve the predictions of the Beckmann-Kirchhoff (BK) model for moderatelyrough surfaces. This paper aims to compare the predictions of a number of alternative models including those of Lafortune et al. (LFTG) and Oren-Nayar (ON) against the BRDF measurements from the CUReT database. Excluding the specular reflectance component, we explore how these models account for the diffuse reflectance component. Our main conclusions are as follows. First, for moderately rough samples Fresnel correction provides the most accurate predictions for the majority of the samples studied. Second, although there are several surface samples in the CUReT database for which the ON model provides the best fit, the ON model and the LFTG model fail to fit the data well for most of the surface samples. Third, for all surface samples studied, the Fresnel corrected variant of the BK model gives more realistic predictions than the empirical reflectance models such as the LFTG model. 1

In this paper we explore whether the Fresnel term can be used to improve the predictions of the Beckmann-Kirchhoff (B-K) model for moderately-rough surfaces. Our arm in developing this model is to describe subsurface scattering effects for surfaces of intermediate roughness. We use the BRDF measurements from the CUReT database to compare the predictions of the Fresnel correction process with several variants of the B-K model and the Oren and Nayar Model. The study reveals that our new Fresnel correction provides accurate predictions, which are considerably better than those achieved using both the alternative variants of the B-K model and the Oren-Nayar model.

In this paper we suggest a modification to the Beckmann-Kirchhoff model of rough surface reflectance that can be used for surfaces of intermediate roughness. One of the problems with the existing Beckmann model is that it fails for large angles of incidence. This problem was overcome for rough surfaces by Vernold and Harvey who use a Lambertian form factor. Here we take this work one step further by using a Fresnel coefficient to account for subsurface reflectance attenuation. This study also aims to show the existing modelling gaps between different variants of the Beckmann model. 1

In this paper we illustrate the use of the Beckmann-Kirchhoff model for analysing rough surface reflectance. The Beckmann-Kirchhoff model is a physical model that describes the reflectance of light from rough surfaces. The parameter of the model is the surface slope, or ratio of the surface roughness to the correlation length. We show how this parameter may be estimated using pairs of surface images, subject to different illumination directions. With the parameter to hand, the Beckmann-Kirchhoff model may be used to perform Lambertian reflectance correction, and hence shape-from-shading may be applied to the rough surfaces. The model may also be used for surface re-illumination. We present experiments to illustrate the utility of the method for each of these tasks. 1

We make use of the Beckmann-Kirchhoff and Davies scattering models to estimate surface properties for both dielectric and metallic surfaces based on reflectance measurements. In the case of metallic surfaces, we consider two refinements of the Davies theory, which apply under different restrictions concerning the reflectance geometry. The first of these is due to Bennett and Porteus (1961) and applies for normal incidence and reflectance. The second is due to Torrance (1964) and applies when the incidence radiation is off normal. We then suggest three classes of materials for which the appropriate approximations may be used to estimate the surface roughness, the correlation length and the surface slope. Finally, we use the surface slope estimates to fit the Beckmann-Kirchhoff model to reflectance data. In contrast to previous methods which work at long wavelengths and use special purpose instrumentation, our methods can be performed using visible light and a digital camera.

For rough and smooth surfaces, there are departures from Lambertian reflectance at the occluding limb. This is due to backscattering from microfacets which protrude above the surface. These departures from Lambertian reflectance limit the applicability of shape-from-shading to the surface. We consider how to reconstruct the Lambertian reflectance component in order to improve the quality of surface normal information recoverable using shape-from-shading. We show that although this correction process is reasonable when the light source and viewing directions are close to one another, it is not applicable when the directions are perpendicular.

We use the viewpoint compensation method of Woodham (1994) to make improved estimates of the Hessian matrix from the Gauss map in shape-from-shading. We use the improved estimates to perform curvature sensitive surface-smoothing. Rather than directly smoothing the surface normal vectors, we exert control over their directions by smoothing the field of principal curvature directions. To do this we develop a topography sensitive smoothing process which overcomes the problems of singularities in the field of principal directions at the locations of umbilics and saddles.

In this paper we address the problem of separating Lambertian and specu-lar reflection components in order to improve the quality of surface normal information recoverable using shape-from-shading. The framework for our study is provided by the iterated conditional modes algorithm. We develop a maximum a posteriori probability (MAP) estimation method for estimat-ing the mixing proportions for Lambertian and specular reflectance, and also, for recovering local surface normals. The MAP estimation scheme has two model ingredients. Firstly, there are separate conditional measurement den-sities which describe the distributions of surface normal directions for the Lambertian and specular reflectance components. The second ingredient is a smoothness prior which models the distribution of surface normal directions over local image regions. We experiment with the method on real-world data. This reveals not only that the method accurately estimates the proportion of specular reflection, but that it also results in good surface normal reconstruc-tion in the proximity of specular highlights. 1

This paper describes a new surface normal smoothing process which can be used in conjunction with shape-from-shading. Rather than directly smoothing the surface normal vectors, we exert control over their directions by smoothing the field of principal curvature vectors. To do this we develop a topography sensitive smoothing process which overcomes the problems of singularities in the field of principal curvature directions at the locations of umbilics and saddles.