Object identification and retrieval from efficient image matching. Snap2Tell with the STOIC dataset

Traditional content based image retrieval attempts to retrieve images using syntactic features for a query image. Annotated image banks and Google allow the use of text to retrieve images. In this paper, we studied the task of using the content of an image to retrieve information in general. We describe the significance of object identification in an information retrieval paradigm that uses image set as intermediate means in indexing and matching. We also describe a unique Singapore Tourist Object Identification Collection with associated queries and relevance judgments for evaluating the new task and the need for efficient image matching using simple image features. We present comprehensive experimental evaluation on the effects of feature dimensions, context, spatial weightings, coverage of image indexes, and query devices on task performance. Lastly we describe the current system developed to support mobile image-based tourist information retrieval.     

Document reranking by term distribution and maximal marginal relevance for chinese information retrieval

In this paper, we propose a document reranking method for Chinese information retrieval. The method is based on a term weighting scheme, which integrates local and global distribution of terms as well as document frequency, document positions and term length. The weight scheme allows randomly setting a larger portion of the retrieved documents as relevance feedback, and lifts off the worry that very fewer relevant documents appear in top retrieved documents. It also helps to improve the performance of maximal marginal relevance (MMR) in document reranking. The method was evaluated by MAP (mean average precision), a recall-oriented measure. Significance tests showed that our method can get significant improvement against standard baselines, and outperform relevant methods consistently.     

Network-based synchronization of T–S fuzzy chaotic systems with asynchronous samplings

In this paper, we study the problem of network-based synchronization of chaotic systems in Takagi–Sugeno (T–S) fuzzy form, in which the master and slave fuzzy chaotic systems are connected with a continuous-time controller through a network. In all communication channels, asynchronous samplings and external disturbances are considered. The asynchronously sampled state information of the master and slave systems received in the controller is treated by designing an observer for estimating the states of the master system. Then, based on the observation result, the problem of asynchronous samplings between the slave-controller and controller-slave channels is solved in two different cases. Sufficient conditions for the existence of the desired observer and controllers for each asynchronous cases are presented in the form of linear matrix inequalities. An numerical example is given to illustrate the validity and potential of the proposed new design techniques.     

The Singapore Football League, 1904–1941: Towards the Institutionalization of Football in the Colonial Port City of Singapore

Abstract

The Singapore Football Association (SFA) was founded in 1892. In 1904, the YMCA initiated the first football league with 12 teams from military and European clubs and School Old Boys’ teams. The first phase from 1904 to 1913 was restricted to European and Eurasian only. The military teams won six out of the nine tournaments. The second phase of the league began in 1917 and from 1921 to 1941. The Straits Chinese Football Association (SCFA) took part in the league and the rejuvenated SFA included a representative from the SCFA. The Singapore Football League started with two divisions 1921 and participating teams from the SCFA in the same year and the Malaya Football Association (MFA) in 1924. The SCFA won the league for the first time in 1925 and subsequently in 1930, 1937, and 1938. In 1929, the SFA was renamed the Singapore Amateur Football Association (SAFA). The MFA won the League for the first time in 1931, and the first local team to win three years in succession from 1931 to 1933. From 1931 to 1941 the local teams won seven league titles out of 11. By 1940 the League grew with 44 teams in three divisions.     

The effect of caffeine ingestion on human neutrophil oxidative burst responses following time-trial cycling

Following fixed-duration exercise of submaximal intensity, caffeine ingestion is associated with an attenuation of the exercise-induced decline in N-formyl-methionyl-phenyl-alanine (f-MLP) stimulated neutrophil oxidative burst. However, the response following high-intensity exhaustive exercise is unknown. Nine endurance-trained male cyclists ingested 6 mg caffeine or placebo per kilogram of body mass 60 min before cycling for 90 min at 70% of maximal oxygen consumption (VO2max) and then performing a time-trial requiring an energy expenditure equivalent to 30 min cycling at 70% maximum power output. Time-trial performance was 4% faster in the caffeine than in the placebo trial (P = 0.043). Caffeine was associated with an increased plasma adrenaline concentration after 90 min of exercise (P = 0.046) and immediately after the time-trial (P = 0.02). Caffeine was also associated with an increased serum caffeine concentration (P < 0.01) after 90 min of exercise and immediately after the time-trial, as well as 1 h after the time-trial. However, the f-MLP-stimulated neutrophil oxidative burst response fell after exercise in both trials (P = 0.002). There was no effect of caffeine on circulating leukocyte or neutrophil counts, but the lymphocyte count was significantly lower on caffeine (20%) after the time-trial (P = 0.003). Our results suggest that high-intensity exhaustive exercise negates the attenuation of the exercise-induced decrease in neutrophil oxidative burst responses previously observed when caffeine is ingested before exercise of fixed duration and intensity. This may be associated with the greater increase in adrenaline concentration observed in the present study.     

Phonological coding and children’s spelling

The goal of this study was to predict children’s spelling from their performance on pseudohomophone choice and rhyme matching tasks. A total of 222 nine- to twelve-year-old children in grades 4, 5, and 6 participated. The children were given, individually, a computerized pseudohomophone choice task with 30 item pairs in two conditions, and a computerized rhyme matching task with 68 word pairs in four conditions. Accuracy and speed of processing were assessed. Three memory tasks, a Spoonerism task, a general ability test, and a written spelling test were also administered. Analyses of variance and multiple regression analyses showed that both accurate and rapid choice of pseudohomophones sounding like real words and rhyme matching contributed substantially to variations in spelling. Orthographic knowledge played a role in spelling as shown by the contribution of word pairs that rhymed but were orthographically dissimilar. Phonological coding as tested by pseudohomophone choice was also important.     

Scan and Learn: Quick Response Code Enabled Museum for Mobile Learning of Anatomy and Pathology

In the past, medical museums played a significant role in anatomy and pathology training. The attraction of medical museums has declined recently due to the emergence of information technology and innovative medical curricula. An innovative mobile learning platform has been developed using quick response (QR) codes for the museum specimens at the Lee Kong Chain School of Medicine, Singapore. High-quality images of the potted specimens were captured and combined into an album and a video using Adobe Acrobat Pro 9 and Windows Movie Maker, respectively. Subsequently, QR codes were generated linking to PDF documents with annotations, pathology, and clinical history concerning the specimens. Quick response codes were piloted in gastrointestinal teaching module for Year 2 medical students. Survey responses were obtained from students to verify the efficacy of QR as a learning tool. The majority of students either agreed or strongly agreed that it was easy to access the information about the specimen with QR codes (4.47 ± 0.84), while 96% of students agreed that they are able to correlate the specimen with the annotated images (4.56 ± 0.56). The majority of students (78%) agreed that QR codes are useful for their learning (4.22 ± 0.87), while 75% of students felt QR codes motivate them to visit Anatomy Resource Centre. Most of the students agreed that QR codes are useful for revision of materials (4.13 ± 1.07) and independent learning (4.38 ± 0.87). These findings suggest that QR codes are not only effective for students learning but also enhance their exploration experience with the museum specimens.