Visual search guided by an efficient top-down attention approach

The selective attention mechanism is employed by humans and primates to realize a truly intelligent perception system, which has the cognitive capability of learning and thinking about how to perceive the environment autonomously. The attention mechanism involves the top-down and bottom-up ways that correspond to the goal-directed and automatic perceptual behaviors, respectively. Rather than considering the automatic perception, this paper presents an artificial system of the goal-directed visual perception by using the object-based top-down visual attention mechanism. This cognitive system can guide the perception to an object of interest according to the current task, context and learned knowledge. It consists of three successive stages: preattentive processing, top-down attentional selection and post-attentive perception. The preattentive processing stage divides the input scene into homogeneous proto-objects, one of which is then selected by the top-down attention and finally sent to the post-attentive perception stage for high-level analysis. Experimental results of target detection in the cluttered environments are shown to validate this system.

We consider three popular types of shared memory that support one of the following sets of operations: {CAS, read, write}, {LL, SC, VL, read, write}, or {RLL, RSC, read, write}. We present algorithms that, together with Moir's [Moi97], efficiently implement each shared memory above from any of the other two. Our implementations are wait-free and have constant time and space complexity. Thus, concurrent programs developed for one of the above memories can be ported to any other without incurring any increase in time complexity. Further, since our implementations are wait-free, a wait-free concurrent program remains wait-free even after porting.

Understanding bottom-up and top-down visual attention mechanisms related to visual quality perception can be greatly beneficial for the design of effective objective quality metrics. Subjective studies based on eye-movement tracking have been recently published that try to get more insight in these interactions. However, it is still not easy to find coherence across their results, also due to the different methodologies adopted to analyze eye-tracking data. In this paper we propose a robust methodology to measure differences between eye-tracking data collected under different experimental conditions. The proposed method takes into account inter-observer variability and content effects, producing results that give an accurate insight in attention variations.

This paper introduces mass estimation--a base modelling mechanism in data mining. It provides the theoretical basis of mass and an efficient method to estimate mass. We show that it solves problems very effectively in tasks such as information retrieval, regression and anomaly detection. The models, which use mass in these three tasks, perform at least as good as and often better than a total of eight state-of-the-art methods in terms of task-specific performance measures. In addition, mass estimation has constant time and space complexities.

Multiscale face recognition in cluttered backgrounds based on visual attention

Gait recognition and understanding systems have shown a wide-ranging application prospect. However, their use of unstructured data from image and video has affected their performance, e.g., they are easily influenced by multi-views, occlusion, clothes, and object carrying conditions. This paper addresses these problems using a realistic 3-dimensional (3D) human structural data and sequential pattern learning framework with top-down attention modulating mechanism based on Hierarchical Temporal Memory (HTM). First, an accurate 2-dimensional (2D) to 3D human body pose and shape semantic parameters estimation method is proposed, which exploits the advantages of an instance-level body parsing model and a virtual dressing method. Second, by using gait semantic folding, the estimated body parameters are encoded using a sparse 2D matrix to construct the structural gait semantic image. In order to achieve time-based gait recognition, an HTM Network is constructed to obtain the sequence-level gait sparse distribution representations (SL-GSDRs). A top-down attention mechanism is introduced to deal with various conditions including multi-views by refining the SL-GSDRs, according to prior knowledge. The proposed gait learning model not only aids gait recognition tasks to overcome the difficulties in real application scenarios but also provides the structured gait semantic images for visual cognition. Experimental analyses on CMU MoBo, CASIA B, TUM-IITKGP, and KY4D datasets show a significant performance gain in terms of accuracy and robustness.

Data — structured, semi-structured and unstructured is growing at an exponential rate. Most organizations are dependent on this huge volume of data to make real time decisions. To provide accurate information to these decision making models, it is crucial that the data is of superior quality. Since most organizations are migrating from relational to NoSQL databases, it is vital to validate data after the migration process. Considering the different structures of the two databases, the process of validation is a formidable task. This paper talks about an approach to validate data between cross platform databases using denormalized schema structures and bloom filters. Based on our experimental results, we have been able to validate huge datasets and pinpoint the exact corrupted records in constant space and linear time complexity up to a desired error probability.

We aim to model the top-down attention of a Convolutional Neural Network (CNN) classifier for generating task-specific attention maps. Inspired by a top-down human visual attention model, we propose a new backpropagation scheme, called Excitation Backprop, to pass along top-down signals downwards in the network hierarchy via a probabilistic Winner-Take-All process. Furthermore, we introduce the concept of contrastive attention to make the top-down attention maps more discriminative. In experiments, we demonstrate the accuracy and generalizability of our method in weakly supervised localization tasks on the MS COCO, PASCAL VOC07 and ImageNet datasets. The usefulness of our method is further validated in the text-to-region association task. On the Flickr30k Entities dataset, we achieve promising performance in phrase localization by leveraging the top-down attention of a CNN model that has been trained on weakly labeled web images.

Strong evidence has shown that visual processing based on selective attention is both data- and knowledge-driven. However, most of the previous work mainly focused on the former. We propose in this paper a new selective attention visual computing model based on both of them. The novelty lies in: (1) A structure variable non-uniform sampling method is proposed to separate visual computing into foveal and peripheral channel. (2) A combination of the bottom-up and the top-down selective attention mechanism based on a two-layered pyramid is proposed. The data-driven bottom-up selective attention includes the sequential extraction of feature maps, conspicuity maps, and interesting maps based on the multi-channel filtering and relaxation process. The knowledge driven top-down selective attention is based on distributed associative memory mapping. (3) A movement control mechanism is also proposed in this paper. Perfectly good experiment results on artificial and real times demonstrate the validity of our model.

A top-down task-dependent model guides attention to likely target locations in cluttered scenes. Here, a novel biologically plausible top-down auditory attention model is presented to model such task-dependent influences on a given task. First, multi-scale features are extracted based on the processing stages in the central auditory system, and converted to low-level auditory "gist" features. These features capture rough information about the overall scene. Then, the top-down model learns the mapping between auditory gist features and the scene categories. The proposed top-down attention model is tested with prominent syllable detection task in speech. When tested on broadcast news-style read speech using the BU Radio News Corpus, the model achieves 85.8% prominence detection accuracy at syllable level. The results compare well to the reported human performance on this task.

Visual search and other attentionally demanding processes are often guided from the top down when a specific task is given (e.g. Wolfe et al. Vision Research 44, 2004). In the simplified stimuli commonly used in visual search experiments, e.g. red and horizontal bars, the selection of potential features that might be biased for is obvious (by design). In a natural setting with real-world objects, the selection of these features is not obvious, and there is some debate which features can be used for top-down guidance, and how a specific task maps to them (Wolfe and Horowitz, Nat. Rev. Neurosci. 2004). Learning to detect objects provides the visual system with an effective set of features suitable for the detection task, and with a mapping from these features to an abstract representation of the object. We suggest a model, in which V4-type features are shared between object detection and top-down attention. As the model familiarizes itself with objects, i.e. it learns to detect them, it acquires a representation for features to solve the detection task. We propose that by cortical feedback connections, top-down processes can re-use these same features to bias attention to locations with higher probability of containing the target object. We propose a model architecture that allows for such processing, and we present a computational implementation of the model that performs visual search in natural scenes for a given object category, e.g. for faces. We compare the performance of our model to pure bottom-up selection as well as to top-down attention using simple features such as hue.

This paper focuses on Dynamic Implementation using Array and Linked list. In the data structure concept, the list plays a major role in the allocation of data where it is a general overview of creating a node in a linked list using an array with the initialization of size. The node contains data and is interconnected to the next node in the sequence via their links which perform the search operation where the space and time complexity are not wasted in these analyses. Linked lists are a great way to store a theoretically infinite amount of data with a small and versatile amount of code. The amount of data stored increases linearly with the number of nodes in the list (space-o(n)) where linked lists are more efficient than arrays. It is observed that considering the time complexity when there is a variation in data set linked array is almost reaches close to array whenever dataset is high also linked array has better time complexity than linked list when the searching element is in middle and last. Considering the space complexity linked array is almost 2 times and 4 times less space is needed than array and linked list respectively.

What is the relationship between top-down and bottom-up attention? Are both types of attention tightly interconnected, or are they independent? We investigated this by testing a large representative sample of the Dutch population on two attentional tasks: a visual search task gauging the efficiency of top-down attention and a singleton capture task gauging bottom-up attention. On both tasks we found typical performance--i.e., participants displayed a significant search slope on the search task and significant slowing caused by the unique, but irrelevant, object on the capture task. Moreover, the high levels of significance we observed indicate that the current set-up provided very high signal to noise ratios, and thus enough power to accurately unveil existing effects. Importantly, in this robust investigation we did not observe any correlation in performance between tasks. The use of Bayesian statistics strongly confirmed that performance on both tasks was uncorrelated. We argue that the current results suggest that there are two attentional systems that operate independently. We hypothesize that this may have implications beyond our understanding of attention. For instance, it may be that attention and consciousness are intertwined differently for top-down attention than for bottom-up attention.

WITHDRAWN: Spearman rank correlation based bagging ensemble bayes optimal classification for genomic predictive pattern analytics

The classic buffer insertion algorithm of van Ginneken has time and space complexity O(n/sup 2/), where n is the number of possible buffer positions. For more than a decade, van Ginneken's algorithm has been the foundation of buffer insertion. In this paper, we present a new algorithm that computes the same optimal buffer insertion, but runs much faster. For 2-pin nets, our time complexity is O(nlogn) and space complexity is O(n). For multipin nets, our time complexity is O(nlog/sup 2/n) and space complexity is O(nlogn). The speedup is achieved by four novel techniques: predictive pruning, candidate tree, fast redundancy check, and fast merging. On industrial test cases, the new algorithms is 2-80 times faster than van Ginneken's algorithm and uses 1/4-1/500 of the memory. Since van Ginneken's algorithm and its variations are used by most existing algorithms on buffer insertion and buffer sizing, our new algorithm significantly improves the performance of all these algorithms. The predictive pruning technique has been applied to buffer cost minimization (Shi et al., 2004), and significantly improved the running time.