Topic 1. Some steric-height features of the water off the Pacific coast of Canada; Topic 2. Possible warming of the water at station P in the northeast Pacific during 1956-1984 [abstract]

TOPIC 1: In terms of seasonal scale, temperature effect dominates the annual change of steric height in the open ocean whereas salinity effect controls it along the continental shelf. Large portion of the annual change of height relative to the 1000-db surface is contained in the upper 100m layer. However, in interannual scale large anomalies of steric height in the open ocean, are more often than not, caused by halosteric rather than thermosteric effect. At least in the open ocean the heights are almost totally determined by the behavior of deep water. Their interannual variability appears to be related to the cumulative effect of Eckman pumping. TOPIC 2: There is a "trend" that over the past 28 years the water at Station P has warmed. Least-square analysis indicates that this warming may be significant but shortening of the time-series data by approximately 10 years fails to show that this is the case. These "trends" have to be interpreted with care. The warming may be "apparent" in that it is not indicated clearly in the deep isopynal surfaces which, during the above period, have deepened. Thus warming at the isobaric surfaces may be the effect of the downward migration of the isopynal surfaces.

On the relative utility of different methods to increase the vertical height in an otter trawl

Comparative fishing experiments to study the relative utility of different methods for increased vertical spread of bottom trawl and the availability of off bottom fishes in the region were made using gussets, kite, separate float line and side panels on a two seam net. The catch rates as well as composition of fish were studied. The opening of the trawl mouth, both horizontally and vertically, under different operating gears and towing tension on warps were measured and estimated for comparison purposes. Better catch rate with good quality fishes was obtained with the gear operated with separate float line. With kite, the vertical spread was increased with less catch indicating poor concentration of off bottom fishes.

Decision tree-based context clustering based on cross validation and hierarchical priors

The standard, ad-hoc stopping criteria used in decision tree-based context clustering are known to be sub-optimal and require parameters to be tuned. This paper proposes a new approach for decision tree-based context clustering based on cross validation and hierarchical priors. Combination of cross validation and hierarchical priors within decision tree-based context clustering offers better model selection and more robust parameter estimation than conventional approaches, with no tuning parameters. Experimental results on HMM-based speech synthesis show that the proposed approach achieved significant improvements in naturalness of synthesized speech over the conventional approaches. © 2011 IEEE.

Height-depth (breadth) relationship of shells in the backwater oyster Crassostrea madrasensin (Preston) of the Cochin harbour

A study of the height-depth relation in the Indian backwater oyster Crassostrea madrasensis (Preston) was carried out. The plot of height against depth showed an exponential trend and a relationship of the form H = ADB. Plot of height against depth also showed larger deviations in height for oysters with greater depth. Analysis showed that variations in height do not result in corresponding variations in depth, particularly in oysters with increased height.

Height-length relation of shells in the Indian backwater oyster Crassostrea madrasensis (Preston) of the Cochin harbour

Height-length relationship in Crassostrea madrasensis (Preston) showed an exponential trend and relation in the form, H=ALᴮ. Deviations of actual values from the mean values consequent to the increase in size were noticed. Height and length approximated in oysters of less than 3.5cm in height resulting in orbicular shape. In oyster of shell height 3.5cm to 8cm, increase in height is faster leading to an oval shape. Above 8cm in height, the oysters become further elongated. Height-length relation is non-linear with an index (B value) of 1.1156. A linear relationship also holds good as the B value is not very much different from unity (H=-2.5424+2.0036L).

Spatio-temporal clustering of probabilistic region trajectories

We propose a novel model for the spatio-temporal clustering of trajectories based on motion, which applies to challenging street-view video sequences of pedestrians captured by a mobile camera. A key contribution of our work is the introduction of novel probabilistic region trajectories, motivated by the non-repeatability of segmentation of frames in a video sequence. Hierarchical image segments are obtained by using a state-of-the-art hierarchical segmentation algorithm, and connected from adjacent frames in a directed acyclic graph. The region trajectories and measures of confidence are extracted from this graph using a dynamic programming-based optimisation. Our second main contribution is a Bayesian framework with a twofold goal: to learn the optimal, in a maximum likelihood sense, Random Forests classifier of motion patterns based on video features, and construct a unique graph from region trajectories of different frames, lengths and hierarchical levels. Finally, we demonstrate the use of Isomap for effective spatio-temporal clustering of the region trajectories of pedestrians. We support our claims with experimental results on new and existing challenging video sequences. © 2011 IEEE.

Clustering user data for user modelling in the GUIDE multi-modal set-top box

Chapter 20 Clustering User Data for User Modelling in the GUIDE Multi-modal Set- top Box PM Langdon and P. Biswas 20.1 ... It utilises advanced user modelling and simulation in conjunction with a single layer interface that permits a ...

MCBoost: Multiple classifier boosting for perceptual co-clustering of images and visual features

We present a new co-clustering problem of images and visual features. The problem involves a set of non-object images in addition to a set of object images and features to be co-clustered. Co-clustering is performed in a way that maximises discrimination of object images from non-object images, thus emphasizing discriminative features. This provides a way of obtaining perceptual joint-clusters of object images and features. We tackle the problem by simultaneously boosting multiple strong classifiers which compete for images by their expertise. Each boosting classifier is an aggregation of weak-learners, i.e. simple visual features. The obtained classifiers are useful for object detection tasks which exhibit multimodalities, e.g. multi-category and multi-view object detection tasks. Experiments on a set of pedestrian images and a face data set demonstrate that the method yields intuitive image clusters with associated features and is much superior to conventional boosting classifiers in object detection tasks.