Showing papers by "International Institute of Information Technology, Hyderabad published in 2018"

ASCERTAIN: Emotion and Personality Recognition Using Commercial Sensors

Abstract: We present ASCERTAIN—a multimodal databa AS e for impli C it p ER sonali T y and A ffect recognit I o N using commercial physiological sensors. To our knowledge, ASCERTAIN is the first database to connect personality traits and emotional states via physiological responses . ASCERTAIN contains big-five personality scales and emotional self-ratings of 58 users along with their Electroencephalogram (EEG), Electrocardiogram (ECG), Galvanic Skin Response (GSR) and facial activity data, recorded using off-the-shelf sensors while viewing affective movie clips. We first examine relationships between users’ affective ratings and personality scales in the context of prior observations, and then study linear and non-linear physiological correlates of emotion and personality. Our analysis suggests that the emotion-personality relationship is better captured by non-linear rather than linear statistics. We finally attempt binary emotion and personality trait recognition using physiological features. Experimental results cumulatively confirm that personality differences are better revealed while comparing user responses to emotionally homogeneous videos, and above-chance recognition is achieved for both affective and personality dimensions.

Design of Secure User Authenticated Key Management Protocol for Generic IoT Networks

Abstract: In recent years, the research in generic Internet of Things (IoT) attracts a lot of practical applications including smart home, smart city, smart grid, industrial Internet, connected healthcare, smart retail, smart supply chain and smart farming. The hierarchical IoT network (HIoTN) is a special kind of the generic IoT network, which is composed of the different nodes, such as the gateway node, cluster head nodes, and sensing nodes organized in a hierarchy. In HIoTN, there is a need, where a user can directly access the real-time data from the sensing nodes for a particular application in generic IoT networking environment. This paper emphasizes on the design of a new secure lightweight three-factor remote user authentication scheme for HIoTNs, called the user authenticated key management protocol (UAKMP). The three factors used in UAKMP are the user smart card, password, and personal biometrics. The security of the scheme is thoroughly analyzed under the formal security in the widely accepted real-or-random model, the informal security as well as the formal security verification using the widely accepted automated validation of Internet security protocols and applications tool. UAKMP offers several functionality features including offline sensing node registration, freely password and biometric update facility, user anonymity, and sensing node anonymity compared to other related existing schemes. In addition, UAKMP is also comparable in computation and communication costs as compared to other existing schemes.

Provably Secure Authenticated Key Agreement Scheme for Smart Grid

Abstract: Due to the rapid development of wireless communication systems, authentication becomes a key security component in smart grid environments. Authentication then plays an important role in the smart grid domain by providing a variety of security services including credentials’ privacy, session-key (SK) security, and secure mutual authentication. In this paper, we analyze the security of a recent relevant work in smart grid, and it is unfortunately not able to deal with SK-security and smart meter secret credentials’ privacy under the widely accepted Canetti–Krawczyk adversary (CK-adversary) model. We then propose a new efficient provably secure authenticated key agreement scheme for smart grid. Through the rigorous formal security analysis, we show that the proposed scheme achieves the well-known security functionalities including smart meter credentials’ privacy and SK-security under the CK-adversary model. The proposed scheme reduces the computation overheads for both smart meters and service providers. Furthermore, the proposed scheme offers more security functionalities as compared to the existing related schemes.

Fake News Detection in Social Networks via Crowd Signals

Abstract: Our work considers leveraging crowd signals for detecting fake news and is motivated by tools recently introduced by Facebook that enable users to flag fake news. By aggregating users' flags, our goal is to select a small subset of news every day, send them to an expert (e.g., via a third-party fact-checking organization), and stop the spread of news identified as fake by an expert. The main objective of our work is to minimize the spread of misinformation by stopping the propagation of fake news in the network. It is especially challenging to achieve this objective as it requires detecting fake news with high-confidence as quickly as possible. We show that in order to leverage users' flags efficiently, it is crucial to learn about users' flagging accuracy. We develop a novel algorithm, DETECTIVE, that performs Bayesian inference for detecting fake news and jointly learns about users' flagging accuracy over time. Our algorithm employs posterior sampling to actively trade off exploitation (selecting news that maximize the objective value at a given epoch) and exploration (selecting news that maximize the value of information towards learning about users' flagging accuracy). We demonstrate the effectiveness of our approach via extensive experiments and show the power of leveraging community signals for fake news detection.

Secure Biometric-Based Authentication Scheme Using Chebyshev Chaotic Map for Multi-Server Environment

Abstract: Multi-server environment is the most common scenario for a large number of enterprise class applications. In this environment, user registration at each server is not recommended. Using multi-server authentication architecture, user can manage authentication to various servers using single identity and password. We introduce a new authentication scheme for multi-server environments using Chebyshev chaotic map. In our scheme, we use the Chebyshev chaotic map and biometric verification along with password verification for authorization and access to various application servers. The proposed scheme is light-weight compared to other related schemes. We only use the Chebyshev chaotic map, cryptographic hash function and symmetric key encryption-decryption in the proposed scheme. Our scheme provides strong authentication, and also supports biometrics & password change phase by a legitimate user at any time locally, and dynamic server addition phase. We perform the formal security verification using the broadly-accepted Automated Validation of Internet Security Protocols and Applications (AVISPA) tool to show that the presented scheme is secure. In addition, we use the formal security analysis using the Burrows-Abadi-Needham (BAN) logic along with random oracle models and prove that our scheme is secure against different known attacks. High security and significantly low computation and communication costs make our scheme is very suitable for multi-server environments as compared to other existing related schemes.

Biometrics-Based Privacy-Preserving User Authentication Scheme for Cloud-Based Industrial Internet of Things Deployment

Abstract: Due to the widespread popularity of Internet-enabled devices, Industrial Internet of Things (IIoT) becomes popular in recent years. However, as the smart devices share the information with each other using an open channel, i.e., Internet, so security and privacy of the shared information remains a paramount concern. There exist some solutions in the literature for preserving security and privacy in IIoT environment. However, due to their heavy computation and communication overheads, these solutions may not be applicable to wide category of applications in IIoT environment. Hence, in this paper, we propose a new biometric-based privacy preserving user authentication (BP2UA) scheme for cloud-based IIoT deployment. BP2UA consists of strong authentication between users and smart devices using preestablished key agreement between smart devices and the gateway node. The formal security analysis of BP2UA using the well-known real-or-random model is provided to prove its session key security. Moreover, an informal security analysis of BP2UA is also given to show its robustness against various types of known attacks. The computation and communication costs of BP2UA in comparison to the other existing schemes of its category demonstrate its effectiveness in the IIoT environment. Finally, the practical demonstration of BP2UA is also done using the NS2 simulation.

Chaotic Map-Based Anonymous User Authentication Scheme With User Biometrics and Fuzzy Extractor for Crowdsourcing Internet of Things

Abstract: The recent proliferation of mobile devices, such as smartphones and wearable devices has given rise to crowdsourcing Internet of Things (IoT) applications. E-healthcare service is one of the important services for the crowdsourcing IoT applications that facilitates remote access or storage of medical server data to the authorized users (for example, doctors, patients, and nurses) via wireless communication. As wireless communication is susceptible to various kinds of threats and attacks, remote user authentication is highly essential for a hazard-free use of these services. In this paper, we aim to propose a new secure three-factor user remote user authentication protocol based on the extended chaotic maps. The three factors involved in the proposed scheme are: 1) smart card; 2) password; and 3) personal biometrics. As the proposed scheme avoids computationally expensive elliptic curve point multiplication or modular exponentiation operation, it is lightweight and efficient. The formal security verification using the widely-accepted verification tool, called the ProVerif 1.93, shows that the presented scheme is secure. In addition, we present the formal security analysis using the both widely accepted real-or-random model and Burrows–Abadi–Needham logic. With the combination of high security and appreciably low communication and computational overheads, our scheme is very much practical for battery limited devices for the healthcare applications as compared to other existing related schemes.

Beyond Pixels: Leveraging Geometry and Shape Cues for Online Multi-Object Tracking

Abstract: This paper introduces geometry and object shape and pose costs for multi-object tracking in urban driving scenarios. Using images from a monocular camera alone, we devise pairwise costs for object tracks, based on several 3D cues such as object pose, shape, and motion. The proposed costs are agnostic to the data association method and can be incorporated into any optimization framework to output the pairwise data associations. These costs are easy to implement, can be computed in real-time, and complement each other to account for possible errors in a tracking-by-detection framework. We perform an extensive analysis of the designed costs and empirically demonstrate consistent improvement over the state-of-the-art under varying conditions that employ a range of object detectors, exhibit a variety in camera and object motions, and, more importantly, are not reliant on the choice of the association framework. We also show that, by using the simplest of associations frameworks (two-frame Hungarian assignment), we surpass the state-of-the-art in multi-object-tracking on road scenes. More qualitative and quantitative results can be found at https://junaidcs032.github.io/Geometry_ObjectShape_MOT/.

Design of Secure and Lightweight Authentication Protocol for Wearable Devices Environment

Abstract: Wearable devices are used in various applications to collect information including step information, sleeping cycles, workout statistics, and health-related information. Due to the nature and richness of the data collected by such devices, it is important to ensure the security of the collected data. This paper presents a new lightweight authentication scheme suitable for wearable device deployment. The scheme allows a user to mutually authenticate his/her wearable device(s) and the mobile terminal (e.g., Android and iOS device) and establish a session key among these devices (worn and carried by the same user) for secure communication between the wearable device and the mobile terminal. The security of the proposed scheme is then demonstrated through the broadly accepted real-or-random model, as well as using the popular formal security verification tool, known as the Automated validation of Internet security protocols and applications. Finally, we present a comparative summary of the proposed scheme in terms of the overheads such as computation and communication costs, security and functionality features of the proposed scheme and related schemes, and also the evaluation findings from the NS2 simulation.

Improving CNN-RNN Hybrid Networks for Handwriting Recognition

Abstract: The success of deep learning based models have centered around recent architectures and the availability of large scale annotated data. In this work, we explore these two factors systematically for improving handwritten recognition for scanned off-line document images. We propose a modified CNN-RNN hybrid architecture with a major focus on effective training using: (i) efficient initialization of network using synthetic data for pretraining, (ii) image normalization for slant correction and (iii) domain specific data transformation and distortion for learning important invariances. We perform a detailed ablation study to analyze the contribution of individual modules and present state of art results for the task of unconstrained line and word recognition on popular datasets such as IAM, RIMES and GW.

Phylogenetic classification of the world's tropical forests

Abstract: Knowledge about the biogeographic affinities of the world’s tropical forests helps to better understand regional differences in forest structure, diversity, composition, and dynamics. Such understanding will enable anticipation of region-specific responses to global environmental change. Modern phylogenies, in combination with broad coverage of species inventory data, now allow for global biogeographic analyses that take species evolutionary distance into account. Here we present a classification of the world’s tropical forests based on their phylogenetic similarity. We identify five principal floristic regions and their floristic relationships: (i) Indo-Pacific, (ii) Subtropical, (iii) African, (iv) American, and (v) Dry forests. Our results do not support the traditional neo- versus paleotropical forest division but instead separate the combined American and African forests from their Indo-Pacific counterparts. We also find indications for the existence of a global dry forest region, with representatives in America, Africa, Madagascar, and India. Additionally, a northern-hemisphere Subtropical forest region was identified with representatives in Asia and America, providing support for a link between Asian and American northern-hemisphere forests.

A Retrospective Analysis of the Fake News Challenge Stance Detection Task

Abstract: The 2017 Fake News Challenge Stage 1 (FNC-1) shared task addressed a stance classification task as a crucial first step towards detecting fake news. To date, there is no in-depth analysis paper to critically discuss FNC-1's experimental setup, reproduce the results, and draw conclusions for next-generation stance classification methods. In this paper, we provide such an in-depth analysis for the three top-performing systems. We first find that FNC-1's proposed evaluation metric favors the majority class, which can be easily classified, and thus overestimates the true discriminative power of the methods. Therefore, we propose a new F1-based metric yielding a changed system ranking. Next, we compare the features and architectures used, which leads to a novel feature-rich stacked LSTM model that performs on par with the best systems, but is superior in predicting minority classes. To understand the methods' ability to generalize, we derive a new dataset and perform both in-domain and cross-domain experiments. Our qualitative and quantitative study helps interpreting the original FNC-1 scores and understand which features help improving performance and why. Our new dataset and all source code used during the reproduction study are publicly available for future research.

Secure Message Communication Protocol Among Vehicles in Smart City

Abstract: Secure messages exchange among different vehicles is one of the most challenging tasks in future smart cities. Any malicious activity has the potential to compromise the confidentiality, integrity, and authenticity of messages exchanged between different vehicles. To ensure secure message communication among the vehicles in a smart city environment, a novel scheme using elliptic curve cryptographic (ECC) technique has been presented in this paper. For this purpose, a two-level authentication key exchange scheme has been designed. In the first level authentication, $\text{CH}$ s are verified by series of messages exchanged between $\text{CH}$ s and the $\text{CA}$ . The verified $\text{CH}$ s are responsible for authentication of vehicles in the second level authentication, followed by exchange of messages between $\text{CH}$ and vehicle. The security analysis using widely accepted Burrows–Abadi–Needham logic, formal security analysis using random oracle model and verification using the widely known automated validation of Internet security protocols and applications (AVISPA) tool, and also the informal security analysis have been done with respect to various types of attacks. Moreover, a comparative analysis of the proposed scheme with existing related schemes reveals that it generates low overhead and latency, and high reliability during messages exchange between vehicles and the $\text{CA}$ .

A Novel Authentication and Key Agreement Scheme for Implantable Medical Devices Deployment

Abstract: Implantable medical devices (IMDs) are man-made devices, which can be implanted in the human body to improve the functioning of various organs. The IMDs monitor and treat physiological condition of the human being (for example, monitoring of blood glucose level by insulin pump). The advancement of information and communication technology enhances the communication capabilities of IMDs. In healthcare applications, after mutual authentication, a user (for example, doctor) can access the health data from the IMDs implanted in a patient's body. However, in this kind of communication environment, there are always security and privacy issues, such as leakage of health data and malfunctioning of IMDs by an unauthorized access. To mitigate these issues, in this paper, we propose a new secure remote user authentication scheme for IMDs communication environment to overcome security and privacy issues in existing schemes. We provide the formal security verification using the widely accepted Automated Validation of Internet Security Protocols and Applications tool. We also provide the informal security analysis of the proposed scheme. The formal security verification and informal security analysis prove that the proposed scheme is secure against known attacks. The practical demonstration of the proposed scheme is performed using the broadly accepted NS2 simulation tool. The computation and communication costs of the proposed scheme are also comparable with the existing schemes. Moreover, the scheme provides additional functionality features, such as anonymity, untraceability, and dynamic implantable medical device addition.

Prediction of vegetation dynamics using NDVI time series data and LSTM

Abstract: Understanding and analyzing the changes in vegetation cover is very important in several aspects including climatic changes, water budget, ecological balance and specially to undertake necessary conservation measures. The concept of neural network has gained much significance in the analysis of vegetation dynamics using remote sensing satellite data. In the current study an attempt has been made to predict the vegetation dynamics using MODIS NDVI time series data sets and long short term memory network, an advanced technique adapted from the artificial neural network. The dataset of 861 NDVI images from January 2000 to June 2016 is used for making the time series. The data is segregated into three sets which comprises of training set (70%), validation set (20%), and testing set (10%). To check the reliability of the experiment we have finalised two different regions after extensive research for investigation. These include different terrains in the Great Nicobar Islands, one region along the coast where vegetation has severe ecological damage due to 2004 Indian Ocean tsunami and the other, an interior region which remained imperturbable during the tsunami. Long short term memory network, an advanced neural network is trained with these NDVI values for both the regions separately to predict the future vegetation dynamics. To measure the accuracy of the LSTM network, root mean square error is calculated. The resulting plots from both the experiments indicate that the long short-term memory neural network follows the series in addition to coinciding with the required time series. Also, an unanticipated change in the trend of the NDVI series were well adapted by the network and was able to predict the future NDVI values with good accuracy maintaining RMSE less than 0.03 without providing any supplementary data. By adopting the prescribed method in the paper, anticipation of vegetation changes can be done accurately much ahead of time and take proactive measures accordingly to safeguard and improve the vegetation in any area.

A Retrospective Analysis of the Fake News Challenge Stance-Detection Task

Abstract: The 2017 Fake News Challenge Stage 1 (FNC-1) shared task addressed a stance classification task as a crucial first step towards detecting fake news. To date, there is no in-depth analysis paper to critically discuss FNC-1’s experimental setup, reproduce the results, and draw conclusions for next-generation stance classification methods. In this paper, we provide such an in-depth analysis for the three top-performing systems. We first find that FNC-1’s proposed evaluation metric favors the majority class, which can be easily classified, and thus overestimates the true discriminative power of the methods. Therefore, we propose a new F1-based metric yielding a changed system ranking. Next, we compare the features and architectures used, which leads to a novel feature-rich stacked LSTM model that performs on par with the best systems, but is superior in predicting minority classes. To understand the methods’ ability to generalize, we derive a new dataset and perform both in-domain and cross-domain experiments. Our qualitative and quantitative study helps interpreting the original FNC-1 scores and understand which features help improving performance and why. Our new dataset and all source code used during the reproduction study are publicly available for future research.

Long-Term Visual Object Tracking Benchmark

Abstract: We propose a new long video dataset (called Track Long and Prosper - TLP) and benchmark for single object tracking. The dataset consists of 50 HD videos from real world scenarios, encompassing a duration of over 400 min (676K frames), making it more than 20 folds larger in average duration per sequence and more than 8 folds larger in terms of total covered duration, as compared to existing generic datasets for visual tracking. The proposed dataset paves a way to suitably assess long term tracking performance and train better deep learning architectures (avoiding/reducing augmentation, which may not reflect real world behaviour). We benchmark the dataset on 17 state of the art trackers and rank them according to tracking accuracy and run time speeds. We further present thorough qualitative and quantitative evaluation highlighting the importance of long term aspect of tracking. Our most interesting observations are (a) existing short sequence benchmarks fail to bring out the inherent differences in tracking algorithms which widen up while tracking on long sequences and (b) the accuracy of trackers abruptly drops on challenging long sequences, suggesting the potential need of research efforts in the direction of long-term tracking. Dataset and tracking results are available at https://amoudgl.github.io/tlp/.

Semi-adversarial Networks: Convolutional Autoencoders for Imparting Privacy to Face Images

Abstract: In this paper, we design and evaluate a convolutional autoencoder that perturbs an input face image to impart privacy to a subject. Specifically, the proposed autoencoder transforms an input face image such that the transformed image can be successfully used for face recognition but not for gender classification. In order to train this autoencoder, we propose a novel training scheme, referred to as semi-adversarial training in this work. The training is facilitated by attaching a semi-adversarial module consisting of an auxiliary gender classifier and an auxiliary face matcher to the autoencoder. The objective function utilized for training this network has three terms: one to ensure that the perturbed image is a realistic face image; another to ensure that the gender attributes of the face are confounded; and a third to ensure that biometric recognition performance due to the perturbed image is not impacted. Extensive experiments confirm the efficacy of the proposed architecture in extending gender privacy to face images.

Overtaking Maneuvers in Simulated Highway Driving using Deep Reinforcement Learning

Abstract: Most methods that attempt to tackle the problem of Autonomous Driving and overtaking usually try to either directly minimize an objective function or iteratively in a Reinforcement Learning like framework to generate motor actions given a set of inputs. We follow a similar trend but train the agent in a way similar to a curriculum learning approach where the agent is first given an easier problem to solve, followed by a harder problem. We use Deep Deterministic Policy Gradients to learn overtaking maneuvers for a car, in presence of multiple other cars, in a simulated highway scenario. The novelty of our approach lies in the training strategy used where we teach the agent to drive in a manner similar to the way humans learn to drive and the fact that our reward function uses only the raw sensor data at the current time step. This method, which resembles a curriculum learning approach is able to learn smooth maneuvers, largely collision free, wherein the agent overtakes all other cars, independent of the track and number of cars in the scene.

SecSVA: Secure Storage, Verification, and Auditing of Big Data in the Cloud Environment

Abstract: With the widespread popularity of Internet-enabled devices, there is an exponential increase in the information sharing among different geographically located smart devices. These smart devices may be heterogeneous in nature and may use different communication protocols for information sharing among themselves. Moreover, the data shared may also change with respect to various Vs (volume, velocity, variety, and value) to categorize it as big data. However, as these devices communicate with each other using an open channel, the Internet, there is a higher chance of information leakage during communication. Most of the existing solutions reported in the literature ignore these facts. Keeping focus on these points, in this article, we propose secure storage, verification, and auditing (SecSVA) of big data in cloud environment. SecSVA includes the following modules: an attribute-based secure data deduplication framework for data storage on the cloud, Kerberos-based identity verification and authentication, and Merkle hash-tree-based trusted third-party auditing on cloud. From the analysis, it is clear that SecSVA can provide secure third party auditing with integrity preservation across multiple domains in the cloud environment.

CalibNet: Geometrically Supervised Extrinsic Calibration using 3D Spatial Transformer Networks

Abstract: 3D LiDARs and 2D cameras are increasingly being used alongside each other in sensor rigs for perception tasks. Before these sensors can be used to gather meaningful data, however, their extrinsics (and intrinsics) need to be accurately calibrated, as the performance of the sensor rig is extremely sensitive to these calibration parameters. A vast majority of existing calibration techniques require significant amounts of data and/or calibration targets and human effort, severely impacting their applicability in large-scale production systems. We address this gap with CalibNet: a geometrically supervised deep network capable of automatically estimating the 6-DoF rigid body transformation between a 3D LiDAR and a 2D camera in real-time. CalibNet alleviates the need for calibration targets, thereby resulting in significant savings in calibration efforts. During training, the network only takes as input a LiDAR point cloud, the corresponding monocular image, and the camera calibration matrix K. At train time, we do not impose direct supervision (i.e., we do not directly regress to the calibration parameters, for example). Instead, we train the network to predict calibration parameters that maximize the geometric and photometric consistency of the input images and point clouds. CalibNet learns to iteratively solve the underlying geometric problem and accurately predicts extrinsic calibration parameters for a wide range of mis-calibrations, without requiring retraining or domain adaptation. The project page is hosted at https://epiception.github.io/CalibNet

Providing Healthcare-as-a-Service Using Fuzzy Rule Based Big Data Analytics in Cloud Computing

Abstract: With advancements in information and communication technology, there is a steep increase in the remote healthcare applications in which patients can get treatment from the remote places also. The data collected about the patients by remote healthcare applications constitute big data because it varies with volume, velocity, variety, veracity, and value. To process such a large collection of heterogeneous data is one of the biggest challenges which requires a specialized approach. To address this challenge, a new fuzzy rule based classifier is presented in this paper with an aim to provide Healthcare-as-a-Service. The proposed scheme is based upon the initial cluster formation, retrieval, and processing of the big data in cloud environment. Then, a fuzzy rule based classifier is designed for efficient decision making for data classification in the proposed scheme. To perform inferencing from the collected data, membership functions are designed for fuzzification and defuzzification processes. The proposed scheme is evaluated on various evaluation metrics, such as average response time, accuracy, computation cost, classification time, and false positive ratio. The results obtained confirm the effectiveness of the proposed scheme with respect to various performance evaluation metrics in cloud computing environment.

CalibNet: Geometrically Supervised Extrinsic Calibration using 3D Spatial Transformer Networks

Abstract: 3D LiDARs and 2D cameras are increasingly being used alongside each other in sensor rigs for perception tasks. Before these sensors can be used to gather meaningful data, however, their extrinsics (and intrinsics) need to be accurately calibrated, as the performance of the sensor rig is extremely sensitive to these calibration parameters. A vast majority of existing calibration techniques require significant amounts of data and/or calibration targets and human effort, severely impacting their applicability in large-scale production systems. We address this gap with CalibNet: a self-supervised deep network capable of automatically estimating the 6-DoF rigid body transformation between a 3D LiDAR and a 2D camera in real-time. CalibNet alleviates the need for calibration targets, thereby resulting in significant savings in calibration efforts. During training, the network only takes as input a LiDAR point cloud, the corresponding monocular image, and the camera calibration matrix K. At train time, we do not impose direct supervision (i.e., we do not directly regress to the calibration parameters, for example). Instead, we train the network to predict calibration parameters that maximize the geometric and photometric consistency of the input images and point clouds. CalibNet learns to iteratively solve the underlying geometric problem and accurately predicts extrinsic calibration parameters for a wide range of mis-calibrations, without requiring retraining or domain adaptation. The project page is hosted at this https URL

Automatic Spelling Correction for Resource-Scarce Languages using Deep Learning

Abstract: Spelling correction is a well-known task in Natural Language Processing (NLP). Automatic spelling correction is important for many NLP applications like web search engines, text summarization, sentiment analysis etc. Most approaches use parallel data of noisy and correct word mappings from different sources as training data for automatic spelling correction. Indic languages are resource-scarce and do not have such parallel data due to low volume of queries and non-existence of such prior implementations. In this paper, we show how to build an automatic spelling corrector for resource-scarce languages. We propose a sequence-to-sequence deep learning model which trains end-to-end. We perform experiments on synthetic datasets created for Indic languages, Hindi and Telugu, by incorporating the spelling mistakes committed at character level. A comparative evaluation shows that our model is competitive with the existing spell checking and correction techniques for Indic languages.

Design of an Anonymity-Preserving Group Formation Based Authentication Protocol in Global Mobility Networks

Abstract: Remote user authentication without compromising user anonymity is an emerging area in the last few years. In this paper, we propose a new anonymity preserving mobile user authentication scheme for the global mobility networks (GLOMONETs). We also propose a new anonymity preserving group formation phase for roaming services in GLOMONETs that meets the extended anonymity requirement without compromising any standard security requirements. We provide the security analysis using the widely-accepted Burrows-Abadi-Needham logic and informal analysis for the proposed scheme to show that it is secure against possible well-known attacks, such as replay, man-in-the-middle, impersonation, privileged-insider, stolen smart card, ephemeral secret leakage, and password guessing attacks. In addition, the formal security verification with the help of the broadly accepted automated validation of internet security protocols and applications software simulation tool is tested on the proposed scheme and the simulation results confirm that the proposed scheme is safe. Moreover, the comparative study of the proposed scheme with other relevant schemes reveals that it performs well as compared to other techniques.

Estimation of rainfall thresholds for landslide occurrences in Kalimpong, India

Abstract: The Indian Himalayan locale has been essentially influenced by the increase in the frequency of landslide events. Out of 0.42 million km2 of India’s landmass prone to landslides, 42% falls in the North East Himalaya, especially Darjeeling and Sikkim. The harm due to landslides is massive, causing loss of life, property and agricultural land, thus initiating a dire need for formulating strategies to minimize its impact. There have been many attempts to establish rainfall thresholds on global, regional and local scales which compare analysis at various levels. Rainfall thresholds anticipate landslide occurrence and help in issuing a warning to civil authorities and the general population. However, empirical relations defining the relationship between landslide occurrences and rainfall events in Kalimpong remain unattended. In this paper, rainfall thresholds for landslide occurrence have been ascertained for Kalimpong area of Darjeeling Himalayas, in the Indian province of West Bengal. A threshold for landslide occurrences which describes intensity–duration threshold was estimated using the power law equation. The relationship for the study area is I = 3.52 D−0.41 (I is rainfall intensity (mm/h) and D is duration (h)). Results show that events with a rainfall intensity of 0.95 mm/h with a duration of 24 h have a high risk of slide initiation in this region. It also demonstrates that for 10- and 20-day antecedent rainfall, an intensity of 88.37 and 133.5 mm is required for landslide occurrence in this region. Such data would help in implementing early warning systems that focus on rainfall thresholds and forecasting rainfall measurements. Rainfall thresholds for landslide initiation in Kalimpong can be enhanced with more precipitation and landslide data as and when available.

Geometric Consistency for Self-Supervised End-to-End Visual Odometry

Abstract: With the success of deep learning based approaches in tackling challenging problems in computer vision, a wide range of deep architectures have recently been proposed for the task of visual odometry (VO) estimation. Most of these proposed solutions rely on supervision, which requires the acquisition of precise ground-truth camera pose information, collected using expensive motion capture systems or high-precision IMU/GPS sensor rigs. In this work, we propose an unsupervised paradigm for deep visual odometry learning. We show that using a noisy teacher, which could be a standard VO pipeline, and by designing a loss term that enforces geometric consistency of the trajectory, we can train accurate deep models for VO that do not require ground-truth labels. We leverage geometry as a self-supervisory signal and propose "Composite Transformation Constraints (CTCs)", that automatically generate supervisory signals for training and enforce geometric consistency in the VO estimate. We also present a method of characterizing the uncertainty in VO estimates thus obtained. To evaluate our VO pipeline, we present exhaustive ablation studies that demonstrate the efficacy of end-to-end, self-supervised methodologies to train deep models for monocular VO. We show that leveraging concepts from geometry and incorporating them into the training of a recurrent neural network results in performance competitive to supervised deep VO methods.