International Institute of Information Technology, Hyderabad

About: International Institute of Information Technology, Hyderabad is a education organization based out in Hyderabad, India. It is known for research contribution in the topics: Computer science & Authentication. The organization has 2048 authors who have published 3677 publications receiving 45319 citations. The organization is also known as: IIIT Hyderabad & International Institute of Information Technology (IIIT).

Nose, eyes and ears: Head pose estimation by locating facial keypoints

Abstract: Monocular head pose estimation requires learning a model that computes the intrinsic Euler angles for pose (yaw, pitch, roll) from an input image of human face. Annotating ground truth head pose angles for images in the wild is difficult and requires ad-hoc fitting procedures (which provides only coarse and approximate annotations). This highlights the need for approaches which can train on data captured in controlled environment and generalize on the images in the wild (with varying appearance and illumination of the face). Most present day deep learning approaches which learn a regression function directly on the input images fail to do so. To this end, we propose to use a higher level representation to regress the head pose while using deep learning architectures. More specifically, we use the uncertainty maps in the form of 2D soft localization heatmap images over five facial keypoints, namely left ear, right ear, left eye, right eye and nose, and pass them through an convolutional neural network to regress the head-pose. We show head pose estimation results on two challenging benchmarks BIWI and AFLW and our approach surpasses the state of the art on both the datasets.

Efficient and Secure Time-Key Based Single Sign-On Authentication for Mobile Devices

Abstract: In recent years, mobile devices are becoming an integrated part of our society, and this reinforces the need for security and privacy without incurring additional communication and computation costs. In this paper, we propose a new efficient privacy preserving time-key-based single sign-on (TK-SSO) authenticated key management protocol for mobile devices using elliptic curve cryptography. This allows us to achieve the desirable security properties along with significantly reduced computation and communication costs. TK-SSO also supports the revocation of mobile users and servers. We prove the security of TK-SSO in a widely accepted adversary real-or-random model, as well as using Burrows–Abadi–Needham (BAN) logic and the Automated Validation of Internet Security Protocols and Applications (AVISPA) simulation tool to demonstrate that TK-SSO can resist various known attacks. We then evaluate the performance of TK-SSO and three related protocols to demonstrate its utility.

Discovering gender differences in facial emotion recognition via implicit behavioral cues

Abstract: We examine the utility of implicit behavioral cues in the form of EEG brain signals and eye movements for gender recognition (GR) and emotion recognition (ER). Specifically, the examined cues are acquired via low-cost, off-the-shelf sensors. We asked 28 viewers (14 female) to recognize emotions from unoccluded (no mask) as well as partially occluded (eye and mouth masked) emotive faces. Obtained experimental results reveal that (a) reliable GR and ER is achievable with EEG and eye features, (b) differential cognitive processing especially for negative emotions is observed for males and females and (c) some of these cognitive differences manifest under partial face occlusion, as typified by the eye and mouth mask conditions.

Electronic structure theory based study of proline interacting with gold nano clusters

Abstract: Interaction between metal nanoparticles and biomolecules is important from the view point of developing and designing biosensors. Studies on proline tagged with gold nanoclusters are reported here using density functional theory (DFT) calculations for its structural, electronic and bonding properties. Geometries of the complexes are optimized using the PBE1PBE functional and mixed basis set, i. e., 6-311++G for the amino acid and SDD for the gold clusters. Equilibrium configurations are analyzed in terms of interaction energies, molecular orbitals and charge density. The complexes associated with cluster composed of an odd number of Au atoms show higher stability. Marked decrease in the HOMO-LUMO gaps is observed on complexation. Major components of interaction between the two moieties are: the anchoring N-Au and O-Au bond; and the non covalent interactions between Au and N-H or O-H bonds. The electron affinities and vertical ionization potentials for all complexes are calculated. They show an increased value of electron affinity and ionization potential on complexation. Natural bond orbital (NBO) analysis reveals a charge transfer between the donor (proline) and acceptor (gold cluster). The results indicate that the nature of interaction between the two moieties is partially covalent. Our results will be useful for further experimental studies and may be important for future applications.

Perfectly Reliable and Secure Communication Tolerating Static and Mobile Mixed Adversary

Abstract: In this paper, we study the problem of perfectly reliable message transmission(PRMT) and perfectly secure message transmission(PSMT) between a sender S and a receiver R in an undirected synchronous network, tolerating a mixed adversary, where the adversary can be either static or mobile. The connectivity requirement, phase complexityand communication complexityare three important parameters of any interactive PRMT/PSMT protocol and are well studied in the literature in the presence of a static/mobile Byzantine adversary. However, in the presence of a mixed adversary, we encounter several surprising consequences. In this paper, we prove that even though the connectivity requirement for PRMT is same against both static and mobile mixed adversary, the lower bound on communication complexity for PRMT tolerating a mobile mixed adversary is more than its static mixed counterpart. This is interesting because against a "Byzantine adversary", the connectivity requirement and the lower bound on the communication complexity of PRMT protocols are same for both static and mobile case. Thus our result shows that for PRMT, a mobile mixed adversary is more powerful than its static counterpart. As our second contribution, we design a four phase communication optimal PSMT protocol tolerating a "static mixed adversary". Comparing this with the existing three phase communication optimal PSMT protocol against a "static Byzantine adversary", we find that additional one phase is enough to design communication optimal protocol against a static mixed adversary. Finally, we show that the connectivity requirement and lower bound on communication complexity of any PSMT protocol is same against both static and mobile mixed adversary, thus proving that mobility of the adversary has no effect on PSMT. To show that our bound is tight, we also present a worst case nine phase communication optimal PSMT protocol tolerating a mobile mixed adversary which is first of it's kind. This also shows that the mobility of the adversary does not hinder to design constant phase communication optimal PSMT protocol. In our protocols, we have used new techniques which can be effectively used against both static and mobile mixed adversary and are of independent interest.