Object segmentation requires both object-level information and low-level pixel data. This presents a challenge for feedforward networks: lower layers in convolutional nets capture rich spatial information, while upper layers encode object-level knowledge but are invariant to factors such as pose and appearance. In this work we propose to augment feedforward nets for object segmentation with a n...

Attention can be focused volitionally by "top-down" signals derived from task demands and automatically by "bottom-up" signals from salient stimuli. The frontal and parietal cortices are involved, but their neural activity has not been directly compared. Therefore, we recorded from them simultaneously in monkeys. Prefrontal neurons reflected the target location first during top-down attention, ...

This paper presents a new methodology for studying different accounts of project management practices based on network mapping and analysis. Drawing upon network mapping and visualization as an analytical strategy top-down and bottom-up accounts of project management practice are analysed and compared. The analysis initially reveals a substantial difference between the top-down and bottom-up ac...

It is solidly established that top-down (goal-driven) and bottom-up (stimulus-driven) attention mechanisms depend on distributed cortical networks, including prefrontal and frontoparietal regions. On the other hand, it is less clear whether the BG also contribute to one or the other of these mechanisms, or to both. The current study was principally undertaken to clarify this issue. Parkinson di...

Top-down and bottom-up theorem proving approaches have each specific advantages and disadvantages. Bottom-up provers profit from strong redundancy control and suffer from the lack of goal-orientation, whereas top-down provers are goal-oriented but have weak calculi when their proof lengths are considered. In order to integrate both approaches our method is to achieve cooperation between a top-d...

Top-down and bottom-up theorem proving approaches have each speciic advantages and disadvantages. Bottom-up provers proot from strong redundancy control and suuer from the lack of goal-orientation, whereas top-down provers are goal-oriented but have weak calculi when their proof lengths are considered. In order to integrate both approaches our method is to achieve cooperation between a top-down...

This thesis investigates two well de ned problems in image segmentation, viz. interactive and semantic image segmentation. Interactive segmentation involves power assisting a user in cutting out objects from an image, whereas semantic segmentation involves partitioning pixels in an image into object categories. We investigate various models and energy formulations for both these problems in thi...

The aim of this study was to clarify the nature of visual processing deficits caused by cerebellar disorders. We studied the performance of two types of visual search (top-down visual scanning and bottom-up visual scanning) in 18 patients with pure cerebellar types of spinocerebellar degeneration (SCA6: 11; SCA31: 7). The gaze fixation position was recorded with an eye-tracking device while the...

Deep learning has been demonstrated to achieve excellent results for image classification and object detection. However, the impact of deep learning on video analysis (e.g. action detection and recognition) has not been that significant due to complexity of video data and lack of annotations. In addition, training deep neural networks on large scale video datasets is extremely computationally e...