Showing papers by "International Institute of Information Technology, Hyderabad published in 2016"

First Person Action Recognition Using Deep Learned Descriptors

Abstract: We focus on the problem of wearer's action recognition in first person a.k.a. egocentric videos. This problem is more challenging than third person activity recognition due to unavailability of wearer's pose and sharp movements in the videos caused by the natural head motion of the wearer. Carefully crafted features based on hands and objects cues for the problem have been shown to be successful for limited targeted datasets. We propose convolutional neural networks (CNNs) for end to end learning and classification of wearer's actions. The proposed network makes use of egocentric cues by capturing hand pose, head motion and saliency map. It is compact. It can also be trained from relatively small number of labeled egocentric videos that are available. We show that the proposed network can generalize and give state of the art performance on various disparate egocentric action datasets.

Towards Sub-Word Level Compositions for Sentiment Analysis of Hindi-English Code Mixed Text

Abstract: Sentiment analysis (SA) using code-mixed data from social media has several applications in opinion mining ranging from customer satisfaction to social campaign analysis in multilingual societies. Advances in this area are impeded by the lack of a suitable annotated dataset. We introduce a Hindi-English (Hi-En) code-mixed dataset for sentiment analysis and perform empirical analysis comparing the suitability and performance of various state-of-the-art SA methods in social media. In this paper, we introduce learning sub-word level representations in our LSTM (Subword-LSTM) architecture instead of character-level or word-level representations. This linguistic prior in our architecture enables us to learn the information about sentiment value of important morphemes. This also seems to work well in highly noisy text containing misspellings as shown in our experiments which is demonstrated in morpheme-level feature maps learned by our model. Also, we hypothesize that encoding this linguistic prior in the Subword-LSTM architecture leads to the superior performance. Our system attains accuracy 4-5% greater than traditional approaches on our dataset, and also outperforms the available system for sentiment analysis in Hi-En code-mixed text by 18%.

A secure and robust temporal credential-based three-factor user authentication scheme for wireless sensor networks

Abstract: User authentication is one of the most important security services required for the resource-constrained wireless sensor networks (WSNs). In user authentication, for critical applications of WSNs, a legitimate user is allowed to query and collect the real-time data at any time from a sensor node of the network as and when he/she demands for it. In order to get the real-time information from the nodes, the user needs to be first authenticated by the nodes as well as the gateway node (GWN) of WSN so that illegal access to nodes do not happen in the network. Recently, Jiang et al. proposed an efficient two-factor user authentication scheme with unlinkability property in WSNs Jiang (2014). In this paper, we analyze Jiang et al.’s scheme. Unfortunately, we point out that Jiang et al.’s scheme has still several drawbacks such as (1) it fails to protect privileged insider attack, (2) inefficient registration phase for the sensor nodes, (3) it fails to provide proper authentication in login and authentication phase, (4) it fails to update properly the new changed password of a user in the password update phase, (5) it lacks of supporting dynamic sensor node addition after initial deployment of nodes in the network, and (6) it lacks the formal security verification. In order to withstand these pitfalls found in Jiang et al.’s scheme, we aim to propose a three-factor user authentication scheme for WSNs. Our scheme preserves the original merits of Jiang et al.’s scheme. Our scheme is efficient as compared to Jiang et al.’s scheme and other schemes. Furthermore, our scheme provides better security features and higher security level than other schemes. In addition, we simulate our scheme for the formal security analysis using the widely-accepted AVISPA (Automated Validation of Internet Security Protocols and Applications) tool. The simulation results clearly demonstrate that our scheme is also secure.

Stochastic Multiple Choice Learning for Training Diverse Deep Ensembles

Abstract: Many practical perception systems exist within larger processes which often include interactions with users or additional components that are capable of evaluating the quality of predicted solutions. In these contexts, it is beneficial to provide these oracle mechanisms with multiple highly likely hypotheses rather than a single prediction. In this work, we pose the task of producing multiple outputs as a learning problem over an ensemble of deep networks -- introducing a novel stochastic gradient descent based approach to minimize the loss with respect to an oracle. Our method is simple to implement, agnostic to both architecture and loss function, and parameter-free. Our approach achieves lower oracle error compared to existing methods on a wide range of tasks and deep architectures. We also show qualitatively that solutions produced from our approach often provide interpretable representations of task ambiguity.

NAPS: Network Analysis of Protein Structures.

Abstract: Traditionally, protein structures have been analysed by the secondary structure architecture and fold arrangement. An alternative approach that has shown promise is modelling proteins as a network of non-covalent interactions between amino acid residues. The network representation of proteins provide a systems approach to topological analysis of complex three-dimensional structures irrespective of secondary structure and fold type and provide insights into structure-function relationship. We have developed a web server for network based analysis of protein structures, NAPS, that facilitates quantitative and qualitative (visual) analysis of residue-residue interactions in: single chains, protein complex, modelled protein structures and trajectories (e.g. from molecular dynamics simulations). The user can specify atom type for network construction, distance range (in A) and minimal amino acid separation along the sequence. NAPS provides users selection of node(s) and its neighbourhood based on centrality measures, physicochemical properties of amino acids or cluster of well-connected residues (k-cliques) for further analysis. Visual analysis of interacting domains and protein chains, and shortest path lengths between pair of residues are additional features that aid in functional analysis. NAPS support various analyses and visualization views for identifying functional residues, provide insight into mechanisms of protein folding, domain-domain and protein-protein interactions for understanding communication within and between proteins. URL:http://bioinf.iiit.ac.in/NAPS/.

Together we stand: Siamese Networks for Similar Question Retrieval

Abstract: Community Question Answering (cQA) services like Yahoo! Answers1, Baidu Zhidao2, Quora3, StackOverflow4 etc. provide a platform for interaction with experts and help users to obtain precise and accurate answers to their questions. The time lag between the user posting a question and receiving its answer could be reduced by retrieving similar historic questions from the cQA archives. The main challenge in this task is the “lexicosyntactic” gap between the current and the previous questions. In this paper, we propose a novel approach called “Siamese Convolutional Neural Network for cQA (SCQA)” to find the semantic similarity between the current and the archived questions. SCQA consist of twin convolutional neural networks with shared parameters and a contrastive loss function joining them. SCQA learns the similarity metric for question-question pairs by leveraging the question-answer pairs available in cQA forum archives. The model projects semantically similar question pairs nearer to each other and dissimilar question pairs farther away from each other in the semantic space. Experiments on large scale reallife “Yahoo! Answers” dataset reveals that SCQA outperforms current state-of-theart approaches based on translation models, topic models and deep neural network https://answers.yahoo.com/ http://zhidao.baidu.com/ http://www.quora.com/ http://stackoverflow.com/ based models which use non-shared parameters.

Provably secure user authentication and key agreement scheme for wireless sensor networks

Abstract: In recent years, user authentication has emerged as an interesting field of research in wireless sensor networks. Most recently, in 2016, Chang and Le presented a scheme to authenticate the users in wireless sensor network using a password and smart card. They proposed two protocols P1 and P2. P1 is based on exclusive or XOR and hash functions, while P2 deploys elliptic curve cryptography in addition to the two functions used in P1. Although their protocols are efficient, we point out that both P1 and P2 are vulnerable to session specific temporary information attack and offline password guessing attack, while P1 is also vulnerable to session key breach attack. In addition, we show that both the protocols P1 and P2 are inefficient in authentication and password change phases. To withstand these weaknesses found in their protocols, we aim to design a new authentication and key agreement scheme using elliptic curve cryptography. Rigorous formal security proofs using the broadly accepted, the random oracle models, and the Burrows-Abadi-Needham logic and verification using the well-known Automated Validation of Internet Security Protocols and Applications tool are preformed on our scheme. The analysis shows that our designed scheme has the ability to resist a number of known attacks comprising those found in both Chang-Le's protocols. Copyright © 2016 John Wiley & Sons, Ltd.

An efficient multi-gateway-based three-factor user authentication and key agreement scheme in hierarchical wireless sensor networks

Abstract: User authentication in wireless sensor network WSN plays a very important role in which a legal registered user is allowed to access the real-time sensing information from the sensor nodes inside WSN. To allow such access, a user needs to be authenticated by the accessed sensor nodes as well as gateway nodes inside WSNs. Because of resource limitations and vulnerability to physical capture of some sensor nodes by an attacker, design of a secure user authentication in WSN continues to be an important and challenging research area in recent years. In this paper, we propose a new three-factor user authentication scheme based on the multi-gateway WSN architecture. Through the widely-accepted Burrows-Abadi-Needham logic, we prove that our scheme provides the secure mutual authentication. We then present the formal security verification of our proposed scheme using AVISPA tool, which is a powerful validation tool for network security applications, and show that our scheme is secure. In addition, the rigorous informal security analysis shows that our scheme is also secure against possible other known attacks including the sensor node capture attack. Furthermore, we present the additional functionality features that our scheme offers, which are efficient in communication and computation. Copyright © 2016 John Wiley & Sons, Ltd.

Deep Feature Embedding for Accurate Recognition and Retrieval of Handwritten Text

Abstract: We propose a deep convolutional feature representation that achieves superior performance for word spotting and recognition for handwritten images. We focus on: -(i) enhancing the discriminative ability of the convolutional features using a reduced feature representation that can scale to large datasets, and (ii) enabling query-by-string by learning a common subspace for image and text using the embedded attribute framework. We present our results on popular datasets such as the IAM corpus and historical document collections from the Bentham and George Washington pages. On the challenging IAM dataset, we achieve a state of the art mAP of 91.58% on word spotting using textual queries and a mean word error rate of 6.69% for the word recognition task.

Design of an efficient and provably secure anonymity preserving three-factor user authentication and key agreement scheme for TMIS

Abstract: Several remote user authentication techniques for telecare medicine information system TMIS have been proposed in the literature. But most existing techniques have limitations such as vulnerable to various attacks, lack of functionalities, and inefficiency. Recently, Amin and Biswas proposed a three-factor authentication and key agreement technique for TMIS. But their scheme is inefficient and has several security drawbacks. The attacks such as privileged-insider, user impersonation, and strong reply attacks are possible on their scheme. It also has flaw in password update phase. In order to overcome drawbacks of their scheme, a new provably secure and efficient three-factor remote user authentication scheme for TMIS is proposed in this paper. The proposed scheme overcomes all drawbacks of their scheme and also provides additional features such as user unlinkability, user anonymity, efficient password, and biometric update. The rigorous informal and formal security analysis using random oracle models and the mostly acceptable Automated Validation of Internet Security Protocols and Applications tool is also performed. During the experimentation, it has been observed that the proposed scheme is secure against various known attacks that include replay and man-in-the-middle attacks. Furthermore, the analysis of computation and communication cost estimation of the proposed scheme depicts that our scheme is efficient as compared with other related exiting schemes. Copyright © 2016 John Wiley & Sons, Ltd.

A Secure Anonymous Authentication Protocol for Mobile Services on Elliptic Curve Cryptography

Abstract: Mobile user authentication is an essential topic to consider in the current communications technology due to greater deployment of handheld devices and advanced technologies. Memon et al. recently proposed an efficient and secure two-factor authentication protocol for location-based services using asymmetric key cryptography. Unlike their claims, the vigilant analysis of this paper substantiates that Memon et al. ’s protocol has quite a few limitations such as vulnerability to key compromised impersonation attack, insecure password changing phase, imperfect mutual authentication, and vulnerability to insider attack. Furthermore, this paper proposes an enhanced secure authentication protocol for roaming services on elliptic curve cryptography. The proposed protocol is also a two-factor authentication protocol and is suitable for practical applications due to the composition of light-weight operations. The proposed protocol’s formal security is verified using Automated Validation of Internet Security Protocols and Applications tool to certify that the proposed protocol is free from security threats. The informal and formal security analyses along with the performance analysis sections determine that the proposed protocol performs better than Memon et al. ’s protocol and other related protocols in terms of security and efficiency.

A secure and efficient ECC-based user anonymity-preserving session initiation authentication protocol using smart card

Abstract: The Session Initiation Protocol (SIP) is a signaling communications protocol, which has been chosen for controlling multimedia communication in 3G mobile networks. The proposed authentication in SIP is HTTP digest based authentication. Recently, Tu et al. presented an improvement of Zhang et al.’s smart card-based authenticated key agreement protocol for SIP. Their scheme efficiently resists password guessing attack. However, in this paper, we analyze the security of Tu et al.’s scheme and demonstrate their scheme is still vulnerable to user’s impersonation attack, server spoofing attack and man-in-the middle attack. We aim to propose an efficient improvement on Tu et al.’s scheme to overcome the weaknesses of their scheme, while retaining the original merits of their scheme. Through the rigorous informal and formal security analysis, we show that our scheme is secure against various known attacks including the attacks found in Tu et al.’s scheme. Furthermore, we simulate our scheme for the formal security analysis using the widely-accepted AVISPA (Automated Validation of Internet Security Protocols and Applications) tool and show that our scheme is secure against passive and active attacks including the replay and man-in-the-middle attacks. Additionally, the proposed scheme is comparable in terms of the communication and computational overheads with Tu et al.’s scheme and other related existing schemes.

SEAP: Secure and efficient authentication protocol for NFC applications using pseudonyms

Abstract: Authentication protocol plays an important role in the short-range wireless communications for the Near Field Communication (NFC) technology. Due to the shared nature of wireless communication networks, there are several kinds of security vulnerabilities. Recently, a pseudonym-based NFC protocol (PBNFCP) has been proposed to withstand the security pitfalls found in the existing conditional privacy preserving security protocol (CPPNFC). However, this paper further analyzes PBNFCP and shows that it still fails to prevent the claimed security properties, such as impersonation attacks against an adversary, who is a malicious registered user having a valid pseudonym and corresponding private key. In order to overcome these security drawbacks, this paper proposes a secure and efficient authentication protocol (SEAP) for NFC applications using lifetime-based pseudonyms. The proposed SEAP is simulated for the formal security verification using the widely-accepted AVISPA (Automated Validation of Internet Security Protocols and Applications) tool. The simulation results show that SEAP is secure. The rigorous security and performance analysis shows that the proposed SEAP is secure and efficient as compared to the related existing authentication protocols for NFC applications.

A novel and provably secure authentication and key agreement scheme with user anonymity for global mobility networks

Abstract: Ubiquitous networks support the roaming service for mobile communication devices. The mobile user can use the services in the foreign network with the help of the home network. Mutual authentication plays an important role in the roaming services, and researchers put their interests on the authentication schemes. Recently, in 2016, Gope and Hwang found that mutual authentication scheme of He et al. for global mobility networks had security disadvantages such as vulnerability to forgery attacks, unfair key agreement, and destitution of user anonymity. Then, they presented an improved scheme. However, we find that the scheme cannot resist the off-line guessing attack and the de-synchronization attack. Also, it lacks strong forward security. Moreover, the session key is known to HA in that scheme. To get over the weaknesses, we propose a new two-factor authentication scheme for global mobility networks. We use formal proof with random oracle model, formal verification with the tool Proverif, and informal analysis to demonstrate the security of the proposed scheme. Compared with some very recent schemes, our scheme is more applicable. Copyright © 2016 John Wiley & Sons, Ltd.

Understanding Language Preference for Expression of Opinion and Sentiment: What do Hindi-English Speakers do on Twitter?

Abstract: Linguistic research on multilingual societies has indicated that there is usually a preferred language for expression of emotion and sentiment (Dewaele, 2010). Paucity of data has limited such studies to participant interviews and speech transcriptions from small groups of speakers. In this paper, we report a study on 430,000 unique tweets from Indian users, specifically Hindi-English bilinguals, to understand the language of preference, if any, for expressing opinion and sentiment. To this end, we develop classifiers for opinion detection in these languages, and further classifying opinionated tweets into positive, negative and neutral sentiments. Our study indicates that Hindi (i.e., the native language) is preferred over English for expression of negative opinion and swearing. As an aside, we explore some common pragmatic functions of code-switching through sentiment detection.

An Efficient Hybrid Anomaly Detection Scheme Using K-Means Clustering for Wireless Sensor Networks

Abstract: Sensor nodes in a wireless sensor network (WSN) may be lost due to enervation or malicious attacks by an adversary. WSNs deployed for several applications including military applications are prone to various attacks, which degrade the network performance very rapidly. Hybrid anomaly is a type of anomaly that contains the different types of attacker nodes such as blackhole, misdirection, wormhole etc. These multiple attacks can be launched in the network using the hybrid anomaly. In this situation, it is very difficult to find out which kind of attacker nodes are activated in the network. This motivates us to design a robust and efficient secure intrusion detection approach in order to extend the lifetime of a WSN. In this paper, we aim to propose a new intrusion detection technique for hybrid anomaly, which uses the existing data mining algorithm, called K-means clustering. For the detection purpose, patterns of intrusions are built automatically by the K-means clustering algorithm over training data. After that intrusions are detected by matching network activities against these detection patterns. We evaluate our approach over a WSN dataset that is created using Opnet modeler, which contains various attributes, such as end-to-end delay, traffic sent and traffic received. The training dataset contains the normal values of the network parameters. The testing dataset is created in actual working mode consists of normal and abnormal values of the network parameters. The proposed technique has the ability to detect two types of malicious nodes: blackhole and misdirection nodes. Our scheme achieves 98.6 % detection rate and 1.2 % false positive rate, which are significantly better than the existing related schemes.

IIIT-CFW: A Benchmark Database of Cartoon Faces in the Wild

Abstract: In this paper, we introduce the cartoon faces in the wild (IIIT-CFW) database and associated problems. This database contains 8,928 annotated images of cartoon faces of 100 public figures. It will be useful in conducting research on spectrum of problems associated with cartoon understanding. Note that to our knowledge, such realistic and large databases of cartoon faces are not available in the literature.

Memory efficient mining of periodic-frequent patterns in transactional databases

Abstract: Periodic-frequent patterns are an important class of regularities which exists in a transactional database. A frequent pattern is called periodic-frequent if it appears at regular intervals in a transactional database. In the literature, a model of periodic-frequent patterns was proposed and pattern growth like approaches to extract patterns are being explored. In these approaches, a periodic-frequent pattern tree is built in which a transaction-id list is maintained at each path's tail-node. As the typical size of transactional database is very huge in the modern e-commerce era, extraction of periodic-frequent patterns by maintaining transaction-ids in the tree requires more memory. In this paper, to reduce the memory requirements, we introduced a notion of period summary by capturing the periodicity of the patterns in a sequence of transaction-ids. While building the tree, the period summary of the transactions is computed and stored at the tail-node of the tree instead of the transaction-ids. We have also proposed a merging framework for period summaries for mining periodic-frequent patterns. The performance could be improved significantly as the memory required to store the period summaries is significantly less than the memory required to store the transaction-id list. Experimental results show that the proposed approach reduces the memory consumption significantly and also improves the runtime efficiency considerably over the existing approaches.

An enhanced and secure trust-extended authentication mechanism for vehicular ad-hoc networks

Abstract: Vehicular Ad-hoc Networks VANETs are a move towards regulating safe traffic and intelligent transportation system. A VANETs is characterized by extremely dynamic topographical conditions owing to speedily moving vehicles. In VANETs, vehicles can transmit messages within a pre-defined area to achieve safety and efficiency of the system. Then ensuring authenticity of origin of messages to the receiver in such a dynamic environment is a crucial challenge. Another concern in VANET is preservation of privacy of user/vehicle. Recently, Chuang and Lee proposed a trust-extended authentication mechanism TEAM for vehicle-to-vehicle communications in VANETs. TEAM not only satisfies various security features but also enhances the performance of the authentication process using transitive trust relationship among vehicles. Nonetheless, our analysis shows that TEAM is vulnerable to insider attack, privacy breach, impersonation attacks and some other problems. In this paper, to eradicate the vulnerabilities found in Chuang-Lee's scheme, an enhanced trust-extended authentication scheme for VANET is proposed. We display the efficiency of our scheme through security analysis and comparison. Through simulation results using widely accepted NS-2 simulator, we show that our scheme authenticates vehicles faster than Chuang-Lee's scheme. Copyright © 2016 John Wiley & Sons, Ltd.

Forest fire risk modeling in Uttarakhand Himalaya using TERRA satellite datasets

Abstract: Forest fire is one of the major causes of degradation in western Himalaya, and is an annual phenomenon in more than 50% in the forests of Uttarakhand state. Fire danger models are useful for the fi...

Monocular reconstruction of vehicles: Combining SLAM with shape priors

Abstract: Reasoning about objects in images and videos using 3D representations is re-emerging as a popular paradigm in computer vision. Specifically, in the context of scene understanding for roads, 3D vehicle detection and tracking from monocular videos still needs a lot of attention to enable practical applications. Current approaches leverage two kinds of information to deal with the vehicle detection and tracking problem: (1) 3D representations (eg. wireframe models or voxel based or CAD models) for diverse vehicle skeletal structures learnt from data, and (2) classifiers trained to detect vehicles or vehicle parts in single images built on top of a basic feature extraction step. In this paper, we propose to extend current approaches in two ways. First, we extend detection to a multiple view setting. We show that leveraging information given by feature or part detectors in multiple images can lead to more accurate detection results than single image detection. Secondly, we show that given multiple images of a vehicle, we can also leverage 3D information from the scene generated using a unique structure from motion algorithm. This helps us localize the vehicle in 3D, and constrain the parameters of optimization for fitting the 3D model to image data. We show results on the KITTI dataset, and demonstrate superior results compared with recent state-of-the-art methods, with upto 14.64 % improvement in localization error.

A DSL Compiler for Accelerating Image Processing Pipelines on FPGAs

Abstract: This paper describes an automatic approach to accelerate image processing pipelines using FPGAs. An image processing pipeline can be viewed as a graph of interconnected stages that processes images successively. Each stage typically performs a point-wise, stencil, or other more complex operations on image pixels. Recent efforts have led to the development of domain-specific languages (DSL) and optimization frameworks for image processing pipelines. In this paper, we develop an approach to map image processing pipelines expressed in the PolyMage DSL to efficient parallel FPGA designs. Our approach exploits reuse and available memory bandwidth (or chip resources) maximally. When compared to Darkroom, a state-of-the-art approach to compile high-level DSL to FPGAs, our approach (a) leads to designs that deliver significantly higher throughput, and (b) supports a greater variety of filters. Furthermore, the designs we generate obtain an improvement even over pre-optimized FPGA implementations provided by vendor libraries for some of the benchmarks.

Glaucoma classification with a fusion of segmentation and image-based features

Abstract: Automated classification of glaucoma is of interest in early detection and treatment. Existing methods employ features which are either image-based or derived from Optic Disc (OD) and Cup (OC) segmentation. While the latter suffers from segmentation inaccuracies, the image-based features tend to overfit in limited availability of training data. We propose a solution to overcome these issues and present a classification framework that fuses both type of features within a co-training based semi-supervised setting to overcome the paucity of labelled data. A novel set of features is proposed to represent the segmented OD-OC regions. Additionally, features based on Texture of projections and color Bag of Visual Words have been designed to be sensitive to the sector-wise deformations in OD. The proposed method was trained on 386 labelled and 717 unlabelled images. It outperformed existing methods with an accuracy and AUC of 73.28%, 0.79 on a private test set of 696 unseen images and 76.77%, 0.78 when cross-tested on DRISHTI-GS1 dataset.

A Novel Chaotic Maps-Based User Authentication and Key Agreement Protocol for Multi-server Environments with Provable Security

Abstract: The widespread popularity of the computer networks has triggered concerns about information security. Password-based user authentication with key agreement protocols have drawn attentions since it provides proper authentication of a user before granting access right to services, and then ensure secure communication over insecure channels. Recently, Lee et al. pointed out different security flaws on Tsaur et al.'s multi-server user authentication protocol, and they further proposed an extended chaotic maps-based user authentication with key agreement protocol for multi-server environments. However, we observed that Lee et al.'s protocol has some functionality and security flaws, i.e., it is inefficient in detection of unauthorized login and it does not support password change mechanism. Besides, their protocol is vulnerable to registration center spoofing attack and server spoofing attack. In order to remedy the aforementioned flaws, we proposed a novel chaotic maps-based user authentication with key agreement protocol for multi-server environments. The proposed protocol is provably secure in the random oracle model under the chaotic-maps based computational Diffie-Hellman assumption. In addition, we analyzed our protocol using BAN logic model. We also compared our protocol with Lee et al.'s protocol in aspects of computation cost, functionalities and securities.