Perceptual Evaluation of Video-Realistic Speech

abstract With many visual speech animation techniques now available, there is a clear need for systematic perceptual evaluation schemes. We describe here our scheme and its application to a new video-realistic (potentially indistinguishable from real recorded video) visual-speech animation system, called Mary 101. Two types of experiments were performed: a) distinguishing visually between real and synthetic image- sequences of the same utterances, ("Turing tests") and b) gauging visual speech recognition by comparing lip-reading performance of the real and synthetic image-sequences of the same utterances ("Intelligibility tests"). Subjects that were presented randomly with either real or synthetic image-sequences could not tell the synthetic from the real sequences above chance level. The same subjects when asked to lip-read the utterances from the same image-sequences recognized speech from real image-sequences significantly better than from synthetic ones. However, performance for both, real and synthetic, were at levels suggested in the literature on lip-reading. We conclude from the two experiments that the animation of Mary 101 is adequate for providing a percept of a talking head. However, additional effort is required to improve the animation for lip-reading purposes like rehabilitation and language learning. In addition, these two tasks could be considered as explicit and implicit perceptual discrimination tasks. In the explicit task (a), each stimulus is classified directly as a synthetic or real image-sequence by detecting a possible difference between the synthetic and the real image-sequences. The implicit perceptual discrimination task (b) consists of a comparison between visual recognition of speech of real and synthetic image-sequences. Our results suggest that implicit perceptual discrimination is a more sensitive method for discrimination between synthetic and real image-sequences than explicit perceptual discrimination.

Desenvolvimento humano, lazer e educação fisica escolar : o papel do componente ludico da cultura

Universidade Estadual de Campinas. Faculdade de Educação Física

Epidemiologia do Padrão Face Longa em escolares do Ensino Fundamental do município de Bauru - SP

OBJETIVO: o objetivo deste trabalho foi classificar e determinar a prevalência dos indivíduos portadores de comprometimento vertical nas relações faciais, conforme a severidade da discrepância e, especialmente, dos portadores de Padrão Face Longa. MÉTODOS: a amostra constou de 5.020 sujeitos de nacionalidade brasileira, de ambos os sexos, com idades entre 10 anos e 16 anos e 11 meses, matriculados no Ensino Fundamental do município de Bauru-SP. O exame da morfologia facial constou da observação direta da face, em normas frontal e lateral, sempre com os lábios em repouso, buscando identificar aqueles indivíduos que apresentassem comprometimento vertical nas relações faciais. Esses, uma vez identificados, foram classificados, considerando-se a severidade, em três subtipos: moderado, médio e grave. Para fins de determinação da prevalência dos portadores de Padrão Face Longa, apenas os classificados como dos subtipos médio e grave foram considerados. RESULTADOS: observou-se uma prevalência de 34,94% de comprometimento vertical nas relações faciais e 14,06% de Padrão Face Longa. CONCLUSÃO: os resultados obtidos no presente estudo permitiram concluir que a prevalência do comprometimento vertical nas relações faciais e do Padrão Face Longa foi maior do que a presumida com base na literatura.

Estudo das características cefalométricas em adolescentes brasileiros portadores de Padrão Face Longa

OBJETIVO: definir valores cefalométricos esqueléticos e dentários para adolescentes brasileiros com Padrão Face Longa. MÉTODOS: a amostra foi constituída de telerradiografias em norma lateral de 30 pacientes com Face Longa, sendo 17 do sexo feminino e 13 do masculino; e 30 pacientes face Padrão I, 15 do sexo masculino e 15 do feminino, no estágio de dentadura permanente durante a adolescência. As características do Padrão Face Longa foram definidas clinicamente, pela análise facial. As seguintes grandezas cefalométricas foram avaliadas: (1) Comportamento sagital das bases apicais (SNA, SNB, ANB, NAP, Co-A, Co-Gn); (2) Comportamento vertical das bases apicais (SN.PP, SN.PM, ângulo goníaco, AFAT, AFAI, AFAM, AFP, AFATperp, AFAIperp); (3) Comportamento dentoalveolar (1-PP, 6-PP, 1-PM, 6-PM, 1.PP, IMPA); e (4) Proporção entre as alturas faciais (AFAIPerp/AFATPerp, AFAI/AFAT, AFAM/AFAI). RESULTADOS E CONCLUSÕES: o erro vertical na Face Longa concentra-se no terço inferior. A maxila apresenta uma maior altura dentoalveolar e a mandíbula, com morfologia mais vertical, mostra maior rotação no sentido horário. Essas características morfológicas e espaciais acarretam alterações sagitais e verticais no esqueleto e alterações verticais dentoalveolares. No sentido sagital, os ângulos de convexidade facial estão aumentados. No sentido vertical, as alturas faciais anteriores total e inferior estão aumentadas. O componente dentoalveolar está mais longo.

A novel polar-based human face recognition computational model

Motivated by a recently proposed biologically inspired face recognition approach, we investigated the relation between human behavior and a computational model based on Fourier-Bessel (FB) spatial patterns. We measured human recognition performance of FB filtered face images using an 8-alternative forced-choice method. Test stimuli were generated by converting the images from the spatial to the FB domain, filtering the resulting coefficients with a band-pass filter, and finally taking the inverse FB transformation of the filtered coefficients. The performance of the computational models was tested using a simulation of the psychophysical experiment. In the FB model, face images were first filtered by simulated V1- type neurons and later analyzed globally for their content of FB components. In general, there was a higher human contrast sensitivity to radially than to angularly filtered images, but both functions peaked at the 11.3-16 frequency interval. The FB-based model presented similar behavior with regard to peak position and relative sensitivity, but had a wider frequency band width and a narrower response range. The response pattern of two alternative models, based on local FB analysis and on raw luminance, strongly diverged from the human behavior patterns. These results suggest that human performance can be constrained by the type of information conveyed by polar patterns, and consequently that humans might use FB-like spatial patterns in face processing.

Influence of physical and geometrical parameters on three-dimensional load transfer mechanism at tunnel face

Three-dimensional discretizations used in numerical analyses of tunnel construction normally include excavation step lengths much shorter than tunnel cross-section dimensions. Simulations have usually worked around this problem by using excavation steps that are much larger than the actual physical steps used in a real tunnel excavation. In contrast, the analyses performed in this study were based on finely discretized meshes capable of reproducing the excavation lengths actually used in tunnels, and the results obtained for internal forces are up to 100% greater than those found in other analyses available in the literature. Whereas most reports conclude that internal forces depend on support delay length alone, this study shows that geometric path dependency (reflected by excavation round length) is very strong, even considering linear elasticity. Moreover, many other solutions found in the literature have also neglected the importance of the relative stiffness between the ground mass and support structure, probably owing to the relatively coarse meshes used in these studies. The analyses presented here show that relative stiffness may account for internal force discrepancies in the order of 60%. A dimensionless expression that takes all these parameters into account is presented as a good approximation for the load transfer mechanism at the tunnel face.

The solution structure of the N-terminal zinc finger of GATA-1 reveals a specific binding face for the transcriptional co-factor FOG

Zinc fingers (ZnFs) are generally regarded as DNA-binding motifs. However, a number of recent reports have implicated particular ZnFs in the mediation of protein-protein interactions. The N-terminal ZnF of GATA-1 (NF) is one such finger, having been shown to interact with a number of other proteins, including the recently discovered transcriptional co-factor FOG. Here we solve the three-dimensional structure of the NF in solution using multidimensional H-1/N-15 NMR spectroscopy, and we use H-1/N-15 spin relation measurements to investigate its backbone dynamics. The structure consists of two distorted beta-hairpins and a single alpha-helix, and is similar to that of the C-terminal ZnF of chicken GATA-1. Comparisons of the NF structure with those of other C-4-type zinc binding motifs, including hormone receptor and LIM domains, also reveal substantial structural homology. Finally, we use the structure to map the spatial locations of NF residues shown by mutagenesis to be essential for FOG binding, and demonstrate that these residues all lie on a single face of the NE Notably, this face is well removed from the putative DNA-binding face of the NE an observation which is suggestive of simultaneous roles for the NF; that is, stabilisation of GATA-1 DNA complexes and recruitment of FOG to GATA-1-controlled promoter regions.

Dysmorphogenesis in psychosis: Quantitative and qualitative measures involving the head and face

In this study, we examined qualitative and quantitative measures involving the head and face in a sample of patients and well controls drawn from the Brisbane Psychosis Study. Patients with psychosis (n=310) and age and sex-matched controls (n=303) were drawn from a defined catchment area. Features assessed involved hair whorls (position, number, and direction), eyes (epicanthus), supraorbital ridge, ears (low set, protrusion, hypoplasia, ear lobe attachment, asymmetry, helix width), and mouth (palate height and shape, palate ridges, furrowed and bifid tongue). Quantitative measures related to skull size (circumference, width and length) selected facial heights and depths. The impact of selected risk factors (place and season of birth, fathers' occupation at time of birth, selfreported pregnancy and birth complications, family history) were examined in the entire group, while the association between age of onset and dysmorphology was assessed within the patient group. Significant group (cases versus controls) differences included: patients had smaller skull bases, smaller facial heights, larger facial depths, lower set and protruding ears, different palate shape and fewer palate ridges. In the entire sample significant associations included: (a) those with positive family history of mental illness bad smaller head circumference, cranial length and facial heights; (b) pregnancy and birth complications was associated with smaller facial beights: (c) larger head circumference was associated with higher ranked fathers' occupations at birth. Within the patient group, age of onset was significantly lower in those with more qualitative anomalies or with larger facial heights. The group differences were not due to outliers or distinct subgroups, suggesting that the factors responsible for the differences may be subtle and widely dispersed in the patient group. The Stanley Foundation supported this project.

How Much Older Do You Get When a Wrinkle Appears on Your Face? Modifying Age Estimates by Number of Wrinkles

The present study investigated the influence of wrinkles on facial age judgments. In Experiment 1, preadolescents, young adults, and middle-aged adults made categorical age judgments for male and female faces. The qualitative (type of wrinkle) and quantitative (density of wrinkles and depth of furrows) contributions of wrinkles were analyzed. Results indicated that the greater the number of wrinkles and the depth of furrows, the older a face was rated. The roles of the gender of the face and the age of the participants were discussed. In Experiment 2, participants performed relative age judgments by comparing pairs of faces. Results revealed that the number of wrinkles had more influence on the perceived facial age than the type of wrinkle. A MDS analysis showed the main dimensions on which participants based their judgments, namely, the number of wrinkles and the depth of furrows. We conclude that the quantitative component is more likely to increase perceived facial age. Nevertheless, other variables, such as the gender of the face and the age of the participants, also seem to be involved in the age estimation process.

Comparability between telephone and face-to-face Structured Clinical Interview for DSM-IV in assessing social anxiety disorder

PURPOSE. This article evaluates the comparability of the telephone and in-person Structured Clinical Interview for DSM-IV (SCID) interviews in assessing patients with social anxiety disorder (SAD) as an independent anxiety diagnosis. DESIGN AND METHODS. One hundred subjects were randomly selected and interviewed with the SCID, once by telephone and once in person (1-3 months later). FINDINGS. The prevalence of SAD assessed with the telephone interviews was 56%, whereas the in-person prevalence was 52%, with no statistically significant difference. The test-retest kappa for the 200 interviews was .84, indication of excellent agreement. PRACTICE IMPLICATIONS. These findings, along with the existing evidence of their validity, should encourage the use of SCID by telephone for SAD diagnostic interviews.