A network of integrate and fire neurons for visual selection

Biological systems have facility to capture salient object(s) in a given scene, but it is still a difficult task to be accomplished by artificial vision systems. In this paper a visual selection mechanism based on the integrate and fire neural network is proposed. The model not only can discriminate objects in a given visual scene, but also can deliver focus of attention to the salient object. Moreover, it processes a combination of relevant features of an input scene, such as intensity, color, orientation, and the contrast of them. In comparison to other visual selection approaches, this model presents several interesting features. It is able to capture attention of objects in complex forms, including those linearly nonseparable. Moreover, computer simulations show that the model produces results similar to those observed in natural vision systems.

Using visual cues of microhabitat traits to find home: the case study of a bromeliad-living jumping spider (Salticidae)

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Green-fluorescent protein facilitates rapid in vivo detection of genetically transformed plant cells

Early detection of plant transformation events is necessary for the rapid establishment and optimization of plant transformation protocols. We have assessed modified versions of the green fluorescent protein (GFP) from Aequorea victoria as early reporters of plant transformation using a dissecting fluorescence microscope with appropriate filters. Gfp-expressing cells from four different plant species (sugarcane, maize, lettuce, and tobacco) were readily distinguished, following either Agrobacterium-mediated or particle bombardment-mediated transformation. The identification of gfp-expressing sugarcane cells allowed for the elimination of a high proportion of non-expressing explants and also enabled visual selection of dividing transgenic cells, an early step in the generation of transgenic organisms. The recovery of transgenic cell clusters was streamlined by the ability to visualize gfp-expressing tissues in vitro.

视觉搜索中注意捕获的控制机制

Theories of Visual search generally differentiate between bottom-up control and top-down control. Bottom-up control occurs when visual selection is determined by the stimulus properties in the search field. Top-down control takes place when observers are able to select those stimuli that are in line with their attentional sets. Pure stimulus-driven capture and contingent capture are two main theories on attentional capture by now, in which, theory of pure capture more emphasize bottom-up control, while theory of contingent capture more emphasize top-down control. Besides those two theories, Perceptual load theory of attention provides completely new perspective to explain attentional capture. The aim of this study is to investigate the mechanism of attentional capture in visual search on the basis of the existing theory of attentional capture and Perceptual load theory of attention. Three aspects of questions were explored in this study, which includes: the modulation role of perceptual load on attentional capture; the influence of search mode on attentional capture; and the influence of stimuli’s spatial and temporal characteristics on attentional capture. The results showed that: (1) Attentional capture was modulated by perceptual load in both conditions in which perceptual load manipulated either by amount of stimuli or similarity of stimuli. (2) Search mode did influence attentional capture, but more important, which was also modulated by perceptual load. (3) The spatial characteristics of congruent and incongruent distractor did influence attentional capture, specifically, the further the distractor from the target, the more interference effect the distractor had on visual search. (4) The temporal characteristics of distractor did influence attentional capture, specifically, the pattern of results from the study in which distractor were presented after the search display, were similar to those from the study in which distractors were presented before the search display. In sum, the results indicated that attentional capture are controlled not only by bottom-up factors, top-down factors but also modulated by available attention resources. These findings contribute to resolve the controversy for mechanism of attentional capture. And the potential application of this research was discussed.

Detecção de polimorfismo em porta-enxertos para citros

Objetivando a verificação da existência de plântulas originadas de embrião zigótico em viveiros comerciais de mudas cítricas, realizou-se o presente trabalho, utilizando a técnica de fAFLP e quatro espécies de porta-enxertos para citros. Verificou-se uma base genética estreita entre as espécies testadas, além de grande variabilidade entre os materiais, independentemente do viveiro em que foram coletados, o que permite concluir que a seleção visual, comumente realizada nos viveiros, é ineficiente.

Identification of haploid maize by flow cytometry, morphological and molecular markers

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Seleção de plantas hospedeiras pela aranha bromelícola Psecas chapoda (Salticidae)

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Concepções e expectativas de crianças e de profissionais sobre o espaço na Educação Infantil

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Unmasking the contribution of low-level features to the guidance of attention

The role of low-level stimulus-driven control in the guidance of overt visual attention has been difficult to establish because low- and high-level visual content are spatially correlated within natural visual stimuli. Here we show that impairment of parietal cortical areas, either permanently by a lesion or reversibly by repetitive transcranial magnetic stimulation (rTMS), leads to fixation of locations with higher values of low-level features as compared to control subjects or in a no-rTMS condition. Moreover, this unmasking of stimulus-driven control crucially depends on the intrahemispheric balance between top-down and bottom-up cortical areas. This result suggests that although in normal behavior high-level features might exert a strong influence, low-level features do contribute to guide visual selection during the exploration of complex natural stimuli.

Music selection using a touch screen interface : effect of auditory and visual feedback on driving and usability

This study aimed to examine the effects on driving, usability and subjective workload of performing music selection tasks using a touch screen interface. Additionally, to explore whether the provision of visual and/or auditory feedback offers any performance and usability benefits. Thirty participants performed music selection tasks with a touch screen interface while driving. The interface provided four forms of feedback: no feedback, auditory feedback, visual feedback, and a combination of auditory and visual feedback. Performance on the music selection tasks significantly increased subjective workload and degraded performance on a range of driving measures including lane keeping variation and number of lane excursions. The provision of any form of feedback on the touch screen interface did not significantly affect driving performance, usability or subjective workload, but was preferred by users over no feedback. Overall, the results suggest that touch screens may not be a suitable input device for navigating scrollable lists.

Investigation into DCT Feature Selection for Visual Lip-Based Biometric Authentication

This paper investigated using lip movements as a behavioural biometric for person authentication. The system was trained, evaluated and tested using the XM2VTS dataset, following the Lausanne Protocol configuration II. Features were selected from the DCT coefficients of the greyscale lip image. This paper investigated the number of DCT coefficients selected, the selection process, and static and dynamic feature combinations. Using a Gaussian Mixture Model - Universal Background Model framework an Equal Error Rate of 2.20% was achieved during evaluation and on an unseen test set a False Acceptance Rate of 1.7% and False Rejection Rate of 3.0% was achieved. This compares favourably with face authentication results on the same dataset whilst not being susceptible to spoofing attacks.