Abstract

In this paper, we address the problem of dance style classification to classify Indian dance or any dance in general. We propose a 3-step deep learning pipeline. First, we extract 14 essential joint locations of the dancer from each video frame, this helps us to derive any body region location within the frame, we use this in the second step which forms the main part of our pipeline. Here, we divide the dancer into regions of important motion in each video frame. We then extract patches centered at these regions. Main discriminative motion is captured in these patches. We stack the features from all such patches of a frame into a single vector and form our hierarchical dance pose descriptor. Finally, in the third step, we build a high level representation of the dance video using the hierarchical descriptors and train it using a Recurrent Neural Network (RNN) for classification. Our novelty also lies in the way we use multiple representations for a single video. This helps us to: (1) Overcome the RNN limitation of learning small sequences over big sequences such as dance; (2) Extract more data from the available dataset for effective deep learning by training multiple representations. Our contributions in this paper are three-folds: (1) We provide a deep learning pipeline for classification of any form of dance; (2) We prove that a segmented representation of a dance video works well with sequence learning techniques for recognition purposes; (3) We extend and refine the ICD dataset and provide a new dataset for evaluation of dance. Our model performs comparable or better in some cases than the state-of-the-art on action recognition benchmarks. Keywords: human activity recognition, dance style recognition, joint-localization, pose-descriptor, CNN, RNN, ICD.

Abstract

Over the last decade, the volume of unannotated user-generated web content has skyrocketed but manually annotating data is costly in terms of time and resources. We leverage the advancements in Machine Learning to reduce these costs. We create a semantically searchable dance database with automatic annotation and retrieval. We use a pose estimation module to retrieve body pose and generate Labanotation over recorded videos. Though generic, it provides an essential application due to large amount of videos available online. Labanotation can be further exploited to generate ontology and is also very relevant for preservation and digitization of such resources. We also propose a semi-automatic annotation model which generates semantic annotations over any video archive using only 2-4 manually annotated clips. We experiment on two publicly available ballet datasets. High-level concepts such as ballet pose and steps are used to make the semantic library. These also act as descriptive meta-tags making the videos retrievable using a semantic text or video query.

Abstract

Edge computing has emerged as an effective solution for delay sensitive IoT applications. In the edge-cloud hierarchy, reliability and fault tolerance are important issues. This paper proposes a novel agent-based reliable and fault-tolerant hierarchical IoT-cloud architecture which can survive the failures of the edge servers. In the proposed architecture, the cloud is distributed over four levels (cloud-fog-mist-dew) based on the processing power and distance from the end IoT device. It makes the whole system reliable by replicating the data on the edge of the network. The proposed system is fault tolerant as it redirects the application on an alternate server. In case of a server’s failure, redirection is done at the best possible level in the hierarchy, based on delay-tolerance of the IoT application. The application keeps on working even if the system fails on any level, on cloud, fog or mist. The solution is proposed using mobile agents (MAs) on servers to share systems’ state and other important information with other agents in the hierarchy, to be used for application redirection in case of server’s failure. The proposed system is simulated using Matlab and results prove its efficiency. Index Terms— Edge Computing, Real-time IoT, Reliability, Fault Tolerance, Fault Localization, Cloud, Mobile Agent.

Abstract

A river water quality management model under average climatic conditions may not be able to account for the extreme risk of low water quality which is more prominent under an increase in river water temperature and altered river flows. A modeling framework is developed to assess the risk of river low water quality extremes by integrating a statistical downscaling model based on Canonical Correlation Analysis, risk quantification model based on Frank Archimedean Copula function and multiple logistic regression model integrated with a river water quality simulation model, QUAL2 K. The results reveal that the combination of predicted decrease in low flows of approximately 57% and increase in maximum river water temperatures of approximately 1.2
C has shown an increase of about 46% in risk of low water quality conditions for the future scenarios along Tunga-Bhadra River, India. The extreme risk of low water quality is observed to increase by 50.6% for the period 2020– 2040 when compared with the current extreme conditions of 4.5% and average risk conditions of about 3% for the period 1988–2005. The study captured the occurrence of extremes of low water quality with evidence of a strong link between climate and water quality impairment events.

Abstract

There have been numerous attempts in explaining the general learning behaviours by various cognitive models. Multiple hypotheses have been put further to qualitatively argue the best-fit model for motor skill acquisition task and its variations. In this context, for a discrete sequence production (DSP) task, one of the most insightful models is Verweys Dual Processor Model (DPM). It largely explains the learning and behavioural phenomenon of skilled discrete key-press sequences without providing any concrete computational basis of reinforcement. Therefore, we propose a quantitative explanation for Verweys DPM hypothesis by experimentally establishing a general computational framework for motor skill learning. We attempt combining the qualitative and quantitative theories based on a best-fit model of the experimental simulations of variations of dual processor models. The fundamental premise of sequential decision making for skill learning is based on interacting model-based (MB) and model-free (MF) reinforcement learning (RL) processes. Our unifying framework shows the proposed idea agrees well to Verweys DPM and Fitts three phases of skill learning. The accuracy of our model can further be validated by its statistical fit with the human-generated data on simple environment tasks like the grid-world. There have been numerous attempts in explaining the general learning behaviours by various cognitive models. Multiple hypotheses have been put further to qualitatively argue the best-fit model for motor skill acquisition task and its variations. In this context, for a discrete sequence production (DSP) task, one of the most insightful models is Verweys Dual Processor Model (DPM). It largely explains the learning and behavioural phenomenon of skilled discrete key-press sequences without providing any concrete computational basis of reinforcement. Therefore, we propose a quantitative explanation for Verweys DPM hypothesis by experimentally establishing a general computational framework for motor skill learning. We attempt combining the qualitative and quantitative theories based on a best-fit model of the experimental simulations of variations of dual processor models. The fundamental premise of sequential decision making for skill learning is based on interacting model-based (MB) and model-free (MF) reinforcement learning (RL) processes. Our unifying framework shows the proposed idea agrees well to Verweys DPM and Fitts three phases of skill learning. The accuracy of our model can further be validated by its statistical fit with the human-generated data on simple environment tasks like the grid-world.

Abstract

This paper explores the effectiveness of deep features for doc- ument image segmentation. The document image segmentation problem is modelled as a pixel labeling task where each pixel in the document im- age is classified into one of the predefined labels such as text, comments, decorations and background. Our method first extracts deep features from superpixels of the document image. Then we learn an svm classifier using these features, and segment the document image. Fisher vector encoded convolutional layer features (fv-cnn) and fully connected layer features (fc-cnn) are used in our study. Experiments validate that our method is effective and yields better results for segmenting document images in comparison to the popular approaches on benchmark hand- written datasets.

Abstract

The popularity and applicability of mobile crowdsensing applications are continuously increasing due to the widespread of mobile devices and their sensing and processing capabilities. However, we need to offer appropriate incentives to the mobile users who contribute their resources and preserve their privacy. Blockchain technologies enable semi-anonymous multi-party interactions and can be utilized in crowdsensing applications to maintain the privacy of the mobile users while ensuring first-rate crowdsensed data. In this work, we propose to use blockchain technologies and smart contracts to orchestrate the interactions between mobile crowdsensing providers and mobile users for the case of spatial crowdsensing, where mobile users need to be at specific locations to perform the tasks. Smart contracts, by operating as processes that are executed on the blockchain, are used to preserve users’ privacy and make payments. Furthermore, for the assignment of the crowdsensing tasks to the mobile users, we design a truthful, cost-optimal auction that minimizes the payments from the crowdsensing providers to the mobile users. Extensive experimental results show that the proposed privacy preserving auction outperforms state-of-the-art proposals regarding cost by ten times for high numbers of mobile users and tasks.

Abstract

Finding whether a graph is k-connected, and the identification of its k-connected components is a fundamental problem in graph theory. For this reason, there have been several algorithms for this problem in both the sequential and parallel settings. Several recent sequential and parallel algorithms fork-connectivity rely on one or more breadth-first traversals of the input graph.While BFS can be made very efficient in a sequential setting,the same cannot be said in the case of parallel environments. A major factor in this difficulty is due to the inherent requirement to use a shared queue, balance work among multiple threads in every round, synchronization, and the like. Optimizing the execution of BFS on many current parallel architectures is therefore quite challenging. For this reason, it can be noticed that the time spent by the current parallel graph connectivity algorithms on BFS operations is usually a significant portion of their overall runtime.In this paper, we study how one can, in the context of algorithms for graph connectivity, mitigate the practical in efficiency of relying on BFS operations in parallel. Our technique suggests that such algorithms may not require a BFS of the input graph but actually can work with a sparse spanning subgraph of the input graph. The incorrectness introduced by not using a BFS spanning tree can then be offset by further post-processing stepson suitably defined small auxiliary graphs. Our experiments on finding the 2, and 3-connectivity of graphs on Nvidia K40c GPUs improve the state-of-the-art on the corresponding problems by a factor 2.2x, and 2.1x respectively

Abstract

The goal of this paper is to obtain the robust transistor sizing of the digital logic circuits by minimizing the leakage power keeping the other performance parameters such as propagation delays and area in the bound. To perform such action, Simulated Annealing based Multi-Objective Optimization algorithm has been implemented and further improved the reduction in leakage power using Gradient Descent (MO-SAGRAD).The algorithms are implemented in python and the performance parameters are computed using HSPICE tool at 45nm, 22nmtechnology nodes with Metal Gate High K PTM model cards. Toensure the unruffled functioning of the cells in the automotive applications, the temperature range has been taken between−55oCto125oCrespectively and the supply voltage range istaken as 0.95V to 1.05V for 45nm and 0.76V to 0.84V for22nm technology node. The Overall reduction in leakage power achieved in the logic cells is up to 58% without any penalty in the critical path delay.

Abstract

Nations have set-up Critical Information Infrastruc-ture (CII) protection centers to provide safety of organizations which are deemed to be critical for the nation. In this context,this paper proposes a framework to identify Critical Information Infrastructures (CIIs). This framework has been adopted to identify the banks which are critical for Indian economy. Previously,Reserve Bank of India, the central bank of the country, has announced three banks as systemically important banks. The proposed approach, in this paper, considers a publicly available data set publishing the number (volume) and the amount (value)of financial transactions in the Indian domestic market involving all the banks. The proposed approach finds the individual ranks of each participating bank based on these two parameters.Finally, rank aggregation method combines these two rank lists(based on the value and the volume) to generate a final rank of the banks. Several rank aggregation methods have been compared in this context.