Comparing segmentation strategies for efficient video passage retrieval

We compare the effect of different text segmentation strategies on speech based passage retrieval of video. Passage retrieval has mainly been studied to improve document retrieval and to enable question answering. In these domains best results were obtained using passages defined by the paragraph structure of the source documents or by using arbitrary overlapping passages. For the retrieval of relevant passages in a video, using speech transcripts, no author defined segmentation is available. We compare retrieval results from 4 different types of segments based on the speech channel of the video: fixed length segments, a sliding window, semantically coherent segments and prosodic segments. We evaluated the methods on the corpus of the MediaEval 2011 Rich Speech Retrieval task. Our main conclusion is that the retrieval results highly depend on the right choice for the segment length. However, results using the segmentation into semantically coherent parts depend much less on the segment length. Especially, the quality of fixed length and sliding window segmentation drops fast when the segment length increases, while quality of the semantically coherent segments is much more stable. Thus, if coherent segments are defined, longer segments can be used and consequently less segments have to be considered at retrieval time.

大脑序列记忆信息的提取

The purpose of this study was to examine the cognitive and neural mechanism underlying the serial position effects using cognitive experiments and ERPs(the event related potentials), for 11 item lists in very short-term and the continuous-distractor paradigm with Chinese character. The results demonstrated that when the length of list was 11 Chinese character, and the presentation time, the item interval and the retention interval was 400ms, the primacy effect and recency effect belong to the associative memory and absolute memory respectively. The retrieval of the item at the primacy part depended mainly on the context cues, but the retrieval of the item at the recency part depended mainly on the memory trace. The same results was concluded in the continuous-distractor paradigm (the presentation time was 1sec, the item interval is 12sec, and the retention interval was 30sec). Cognitive results revealed the robust serial position effects in the continuous-distractor paradigm. The different retrieval process between items at the primacy part and items at the recency part of the serial position curve was found. The behavioral responses data of ERP illustrated that the responses for the prime and recent items differed neither in accuracy nor reaction time, the retrieval time for the items at the primacy part was longer than that for the items at the recency part. And the accuracy of retrieval for the primacy part item was lower than that for the recency part items. That meant the retrieval of primacy part items needed more cognitive processes. The recent items, compared with the prime items, evoked ERPs that were more positive, this enhanced positivity occurred in a positive component peaking around 360ms. And for the same retrieval direction (forward or backward), the significant positive component difference between the retrieval for prime items and the retrieval for recent items was found. But there was no significant difference between the forward and backward retrieval at both the primacy and recency part of the serial position curve. These revealed the two kind of retrieval (forward and backward) at the same part of the serial position curve belonged to the same property. These findings fit more closely with the notion of the distinct between the associative memory and the absolute memory.

BoostMap: A Method for Efficient Approximate Similarity Rankings

This paper introduces BoostMap, a method that can significantly reduce retrieval time in image and video database systems that employ computationally expensive distance measures, metric or non-metric. Database and query objects are embedded into a Euclidean space, in which similarities can be rapidly measured using a weighted Manhattan distance. Embedding construction is formulated as a machine learning task, where AdaBoost is used to combine many simple, 1D embeddings into a multidimensional embedding that preserves a significant amount of the proximity structure in the original space. Performance is evaluated in a hand pose estimation system, and a dynamic gesture recognition system, where the proposed method is used to retrieve approximate nearest neighbors under expensive image and video similarity measures. In both systems, BoostMap significantly increases efficiency, with minimal losses in accuracy. Moreover, the experiments indicate that BoostMap compares favorably with existing embedding methods that have been employed in computer vision and database applications, i.e., FastMap and Bourgain embeddings.

Learning Euclidean Embeddings for Indexing and Classification

BoostMap is a recently proposed method for efficient approximate nearest neighbor retrieval in arbitrary non-Euclidean spaces with computationally expensive and possibly non-metric distance measures. Database and query objects are embedded into a Euclidean space, in which similarities can be rapidly measured using a weighted Manhattan distance. The key idea is formulating embedding construction as a machine learning task, where AdaBoost is used to combine simple, 1D embeddings into a multidimensional embedding that preserves a large amount of the proximity structure of the original space. This paper demonstrates that, using the machine learning formulation of BoostMap, we can optimize embeddings for indexing and classification, in ways that are not possible with existing alternatives for constructive embeddings, and without additional costs in retrieval time. First, we show how to construct embeddings that are query-sensitive, in the sense that they yield a different distance measure for different queries, so as to improve nearest neighbor retrieval accuracy for each query. Second, we show how to optimize embeddings for nearest neighbor classification tasks, by tuning them to approximate a parameter space distance measure, instead of the original feature-based distance measure.

A recuperação da informação da memória de trabalho: evidências baseadas em tarefas de imagem mental e de memória visual

A imagem mental e a memória visual têm sido consideradas como componentes distintos na codificação da informação, e associados a processos diferentes da memória de trabalho. Evidências experimentais mostram, por exemplo, que o desempenho em tarefas de memória baseadas na geração de imagem mentais (imaginação visual) sofre a interferência do ruído visual dinâmico (RVD), mas não se observa o mesmo efeito em tarefas de memória visual baseadas na percepção visual (memória visual). Embora várias evidências mostrem que tarefas de imaginação e de memória visual sejam baseadas em processos cognitivos diferentes, isso não descarta a possibilidade de utilizarem também processos em comum e que alguns resultados experimentais que apontam diferenças entre as duas tarefas resultem de diferenças metodológicas entre os paradigmas utilizados para estuda-las. Nosso objetivo foi equiparar as tarefas de imagem mental visual e memória visual por meio de tarefas de reconhecimento, com o paradigma de dicas retroativas espaciais. Sequências de letras romanas na forma visual (tarefa de memória visual) e acústicas (tarefa de imagem mental visual) foram apresentadas em quatro localizações espaciais diferentes. No primeiro e segundo experimento analisou-se o tempo do curso de recuperação tanto para o processo de imagem quanto para o processo de memória. No terceiro experimento, comparou-se a estrutura das representações dos dois componentes, por meio da apresentação do RVD durante a etapa de geração e recuperação. Nossos resultados mostram que não há diferenças no armazenamento da informação visual durante o período proposto, porém o RVD afeta a eficiência do processo de recuperação, isto é o tempo de resposta, sendo a representação da imagem mental visual mais suscetível ao ruído. No entanto, o processo temporal da recuperação é diferente para os dois componentes, principalmente para imaginação que requer mais tempo para recuperar a informação do que a memória. Os dados corroboram a relevância do paradigma de dicas retroativas que indica que a atenção espacial é requisitada em representações de organização espacial, independente se são visualizadas ou imaginadas.

Retrieval of Electron Return Time from High-order Harmonics Generated in a Mixture of He and Ne Gases

In terms of single-atom induced dipole moment by Lewenstein model, we present the macroscopic high-order harmonic generation from mixed He and Ne gases with different mixture ratios by solving three-dimensional Maxwell's equation of harmonic field. And then we show the validity of mixture formulation by Wagner et al. [Phys. Rev. A 76 (2007) 061403(R)] in macroscopic response level. Finally, using least squares fitting we retrieve the electron return time of short trajectory by formulation in Kanai et al. [Phys. Rev. Lett. 98 (2007) 153904] when the gas jet is put after the laser focus.

The spacing effect depends on an encoding deficit, retrieval, and time in working memory: evidence from once-presented words.

The spacing effect in list learning occurs because identical massed items suffer encoding deficits and because spaced items benefit from retrieval and increased time in working memory. Requiring the retrieval of identical items produced a spacing effect for recall and recognition, both for intentional and incidental learning. Not requiring retrieval produced spacing only for intentional learning because intentional learning encourages retrieval. Once-presented words provided baselines for these effects. Next, massed and spaced word pairs were judged for matches on their first three letters, forcing retrieval. The words were not identical, so there was no encoding deficit. Retrieval could and did cause spacing only for the first word of each pair; time in working memory, only for the second.

Real-time sea surface air temperature retrieval over Southern Ocean using AMSR-E data

The AMSR-E satellite data and in-situ data were applied to retrieve sea surface air temperature (Ta) over the Southern Ocean. The in-situ data were obtained from the 24~(th) -26~(th) Chinese Antarctic Expeditions during 2008-2010. First, Ta was used to analyze the relativity with the bright temperature (Tb) from the twelve channels of AMSR-E, and no high relativity was found between Ta and Tb from any of the channels. The highest relativity was 0.38 (with 23.8 GHz). The dataset for the modeling was obtained by using in-situ data to match up with Tb, and two methods were applied to build the retrieval model. In multi-parameters regression method, the Tbs from 12 channels were used to the model and the region was divided into two parts according to the latitude of 50°S. The retrieval results were compared with the in-situ data. The Root Mean Square Error (RMS) and relativity of high latitude zone were 0.96℃and 0.93, respectively. And those of low latitude zone were 1.29 ℃ and 0.96, respectively. Artificial neural network (ANN) method was applied to retrieve Ta.The RMS and relativity were 1.26 ℃ and 0.98, respectively.

Efficient and accurate retrieval of business process models through indexing

Recent years have seen an increased uptake of business process management technology in industries. This has resulted in organizations trying to manage large collections of business process models. One of the challenges facing these organizations concerns the retrieval of models from large business process model repositories. For example, in some cases new process models may be derived from existing models, thus finding these models and adapting them may be more effective than developing them from scratch. As process model repositories may be large, query evaluation may be time consuming. Hence, we investigate the use of indexes to speed up this evaluation process. Experiments are conducted to demonstrate that our proposal achieves a significant reduction in query evaluation time.

Elements pour adapter les systemes de recherche d'information aux dyslexiques [Elements to adapt information retrieval systems to dyslexic users]

RÉSUMÉ. La prise en compte des troubles de la communication dans l’utilisation des systèmes de recherche d’information tels qu’on peut en trouver sur le Web est généralement réalisée par des interfaces utilisant des modalités n’impliquant pas la lecture et l’écriture. Peu d’applications existent pour aider l’utilisateur en difficulté dans la modalité textuelle. Nous proposons la prise en compte de la conscience phonologique pour assister l’utilisateur en difficulté d’écriture de requêtes (dysorthographie) ou de lecture de documents (dyslexie). En premier lieu un système de réécriture et d’interprétation des requêtes entrées au clavier par l’utilisateur est proposé : en s’appuyant sur les causes de la dysorthographie et sur les exemples à notre disposition, il est apparu qu’un système combinant une approche éditoriale (type correcteur orthographique) et une approche orale (système de transcription automatique) était plus approprié. En second lieu une méthode d’apprentissage automatique utilise des critères spécifiques , tels que la cohésion grapho-phonémique, pour estimer la lisibilité d’une phrase, puis d’un texte. ABSTRACT. Most applications intend to help disabled users in the information retrieval process by proposing non-textual modalities. This paper introduces specific parameters linked to phonological awareness in the textual modality. This will enhance the ability of systems to deal with orthographic issues and with the adaptation of results to the reader when for example the reader is dyslexic. We propose a phonology based sentence level rewriting system that combines spelling correction, speech synthesis and automatic speech recognition. This has been evaluated on a corpus of questions we get from dyslexic children. We propose a specific sentence readability measure that involves phonetic parameters such as grapho-phonemic cohesion. This has been learned on a corpus of reading time of sentences read by dyslexic children.

Personalised search – a hybrid approach for web information retrieval and its evaluation

Nowadays, everyone can effortlessly access a range of information on the World Wide Web (WWW). As information resources on the web continue to grow tremendously, it becomes progressively more difficult to meet high expectations of users and find relevant information. Although existing search engine technologies can find valuable information, however, they suffer from the problems of information overload and information mismatch. This paper presents a hybrid Web Information Retrieval approach allowing personalised search using ontology, user profile and collaborative filtering. This approach finds the context of user query with least user’s involvement, using ontology. Simultaneously, this approach uses time-based automatic user profile updating with user’s changing behaviour. Subsequently, this approach uses recommendations from similar users using collaborative filtering technique. The proposed method is evaluated with the FIRE 2010 dataset and manually generated dataset. Empirical analysis reveals that Precision, Recall and F-Score of most of the queries for many users are improved with proposed method.

Secured e-health data retrieval in DaaS and Big Data

Big Data is a rising IT trend similar to cloud computing, social networking or ubiquitous computing. Big Data can offer beneficial scenarios in the e-health arena. However, one of the scenarios can be that Big Data needs to be kept secured for a long period of time in order to gain its benefits such as finding cures for infectious diseases and protecting patient privacy. From this connection, it is beneficial to analyse Big Data to make meaningful information while the data is stored securely. Therefore, the analysis of various database encryption techniques is essential. In this study, we simulated 3 types of technical environments, namely, Plain-text, Microsoft Built-in Encryption, and custom Advanced Encryption Standard, using Bucket Index in Data-as-a-Service. The results showed that custom AES-DaaS has a faster range query response time than MS built-in encryption. Furthermore, while carrying out the scalability test, we acknowledged that there are performance thresholds depending on physical IT resources. Therefore, for the purpose of efficient Big Data management in eHealth it is noteworthy to examine their scalability limits as well even if it is under a cloud computing environment. In addition, when designing an e-health database, both patient privacy and system performance needs to be dealt as top priorities.