Department of Computer Science & Engineering

&ldquo;IIT Jodhpur is committed to impart quality education to facilitate the creation of qualified, competent and responsible human resources to meet the emerging technological challenges of the world. It is working towards creating an equitable learning platform that is positive, accessible and effective in the face of the modern societal and scientific advancements.

Flexibility in curriculum, in all its academic programs, encompassing elements of teaching, research, creativity, innovation and entrepreneurship is the most unique feature of IIT Jodhpur. The Department of Science and Technology (DST), under the National Mission on Cyber Physical Systems, has identified IIT Jodhpur as a Technology &amp; Innovation Hub in the area of Augmented Reality and Virtual Reality. The institute has also been nominated as the coordinator for the Jodhpur City and Knowledge Cluster by the Office of Principal Scientific Adviser, Government of India.

The sprawling 852-acre IIT Jodhpur campus is a majestic tribute to the deep-rooted diverse cultural heritage of the region. The campus is an architectural masterpiece, which is slated to become an international exemplar of sustainability with net-zero energy, and advanced water and waste management strategies.&rdquo;

IIT Jodhpur

The latent fingerprint examiners often mark minutiae and perform a comparison of latent fingerprints with exemplar fingerprints of known identity. The interpretation of details in the fingerprints is often based on the proficiency of examiners. Different examiners discern fingerprint regions and details differently due to an unconscious choice of certain features driven by their experiences. In this study, we aim to draw inferences from the perceptual behavior by collecting eye gaze of examiners while they mark minutiae and perform comparison. The study shows the patterns observed across different forensic examiners and infers specific heuristics used by examiners to discern features. These practices could be inculcated back into an AFIS system to improve automated comparison and help train novice examiners. To draw inferences, novice and expert examiners perform latent to exemplar fingerprint comparison by following the ACE-V mechanism. During the comparison, examiners provide a value determination, quality score and minutiae markup with respective confidences. 29 distinct examiners perform a total of 158 trials, where, the eye gaze is recorded simultaneously. Using the eye gaze fixation, we empirically find Regions of Interest (ROI) of examiners on the prints and utilize it towards developing an understanding of the search strategy.

World University root node

IEEE Transactions on Biometrics, Behavior and Identity Science

Understanding ACE-V Latent Fingerprint Examination Process via Eye-Gaze Analysis

Presentation attacks on face recognition systems are classified into two categories: physical and digital. While much research has focused on physical attacks such as photo, replay and mask attacks, digital attacks such as morphing have received limited attention. With the advancements in deep learning and computer vision algorithms, several easy-to-use applications are available where with few taps/clicks, an image can be easily and seamlessly altered. Moreover, generation of synthetic images or modifying images/videos (e.g. creating deepfakes) is relatively easy and highly effective due to the tremendous improvement in generative machine learning models. Many of these techniques can be used to attack the face recognition systems. To address this potential security risk, in this research, we present a novel algorithm for digital presentation attack detection, termed as MagNet, using a “Weighted Local Magnitude Pattern” (WLMP) feature descriptor. We also present a database, termed as IDAgender, which consists of three different subsets of swapping/morphing and neural face transformation. In contrast to existing research, which utilizes sophisticated machine learning networks for attack generation, the databases in this research are prepared using social media platforms that are readily available to everyone with and without any malicious intent. Experiments on the proposed database, FaceForensic database, GAN generated images and real-world images/videos show the stimulating performance of the proposed algorithm. Through the extensive experiments, it is observed that the proposed algorithm not only yields lower error rates, but also provides computational efficiency.

Fulltext

Frontiers in Artificial Intelligence

MagNet: Detecting Digital Presentation Attacks on Face Recognition

In the near future, one of the promising concepts that are going to change undoubtedly the whole automobile industry, vehicular networking, and transportation sector is the Internet of Vehicles (IoVs). IoVs open new opportunities in different services and applications to enhance road conditions, driver/passenger safety, transportation issues, non-safety applications, and especially for the avoidance of accidents. For reducing accidents, timely warning messages play an important role. For that, lightweight protocols are needed. Another important criterion to be ensured is ‘security’. Because the attackers always try to convey false warning messages that may result in serious issues or attacks. However, in the existing research, secure and lightweight protocols in IoVs are missing. Hence, in this paper, we designed a protocol, P2-SHARP (Privacy-Preserving Secure Hash-based Authentication and Revelation Protocol) using IoVs. In this scheme, secure transmission of local and global warning messages is done. For ensuring security, we have used several cryptographic operations. Security and performance analysis reveals the efficiency of P2-SHARP. © 2021 Elsevier B.V.

Computer Networks

P2-SHARP: Privacy Preserving Secure Hash based Authentication and Revelation Protocol in IoVs

With the advent of new technologies in both hardware and software, we are in the need of a new type of application that requires huge computation power and minimal delay. Applications such as face recognition, augmented reality, virtual reality, automated vehicles, industrial IoT, etc. belong to this category. Cloud computing technology is one of the candidates to satisfy the computation requirement of resource-intensive applications running in UEs (User Equipment) as it has ample computational capacity, but the latency requirement for these applications cannot be satisfied by the cloud due to the propagation delay between UEs and the cloud. To solve the latency issues for the delay-sensitive applications a new network paradigm has emerged recently known as Multi-Access Edge Computing (MEC) (also known as mobile edge computing) in which computation can be done at the network edge of UE devices. To execute the resource-intensive tasks of UEs in the MEC servers hosted in the network edge, a UE device has to offload some of the tasks to MEC servers. Few survey papers talk about task offloading in MEC, but most of them do not have in-depth analysis and classification exclusive to MEC task offloading. In this paper, we are providing a comprehensive survey on the task offloading scheme for MEC proposed by many researchers. We will also discuss issues, challenges, and future research direction in the area of task offloading to MEC servers. © 2021 Elsevier B.V.

Journal of Systems Architecture

A Survey on Task Offloading in Multi-access Edge Computing

Recently, Hosoyamada and Sasaki (EUROCRYPT 2020), and Xiaoyang Dong et al. (ASIACRYPT 2020) proposed quantum collision attacks against AES-like hashing modes AES-MMO and AES-MP. Their collision attacks are based on the quantum version of the rebound attack technique exploiting the differential trails whose probabilities are too low to be useful in the classical setting but large enough in the quantum setting. In this work, we present dedicated quantum free-start collision attacks on Hirose’s double block length compression function instantiated with AES-256, namely HCF-AES-256. The best publicly known classical attack against HCF-AES-256 covers up to 9 out of 14 rounds. We present a new 10-round differential trail for HCF-AES-256 with probability 2−160, and use it to find collisions with a quantum version of the rebound attack. Our attack succeeds with a time complexity of 285.11 and requires 216 qRAM in the quantum-attack setting, where an attacker can make only classical queries to the oracle and perform offline computations. We also present a quantum free-start collision attack on HCF-AES-256 with a time complexity of 286.07 which outperforms Chailloux, Naya-Plasencia, and Schrottenloher’s generic quantum collision attack (ASIACRYPT 2017) in a model when large qRAM is not available. © 2021, Ruhr-Universitat Bochum. All rights reserved.

IACR Transactions on Symmetric Cryptology

Quantum free-start collision attacks on double block length hashing with round-reduced aes-256

We know that TREE CONTRACTION does not admit a polynomial kernel unless NP ⊆ coNP/poly, while PATH CONTRACTION admits a kernel with O(k) vertices. The starting point of this article is the following natural questions: What is the structure of the family of paths that allows PATH CONTRACTION to admit a polynomial kernel? Apart from the size of the solution, what other additional parameters should we consider so we can design polynomial kernels for these basic contraction problems? To design polynomial kernels, we consider the family of trees with the bounded number of leaves (note that the family of paths are trees with at most two leaves). In particular, we study BOUNDED TREE CONTRACTION. Here, an input is a graph G, integers k and ℓ, and the goal is to decide whether, there is a subset F⊆E(G) of size at most k such that G/F is a tree with at most ℓ leaves. We design a kernel with O(kℓ) vertices and O(k2+kℓ) edges for this problem. We complement this result by giving kernelization lower bound. We also prove similar results for BOUNDED OUT-TREE CONTRACTION and BOUNDED CACTUS CONTRACTION. © 2021 Elsevier B.V.

Theoretical Computer Science

Paths to trees and cacti

Pattern Recognition

Context Extraction Module for Deep Convolutional Neural Networks

To increase the recovery of fertilizer nitrogen, reduction in the leaching loss of nitrate-N became a major challenge in nitrogen management in soil. A greenhouse experiment was conducted with maize to study the effect of dose of neem-leaf dust on transformation of nitrogen and the period of efficacy as nitrification inhibitor in a Vertic Epiaquept. The treatments comprised of neem-leaf dust @ 0, 500, 750 and 1000 mg kg−1 of soil at three frequencies, e.g., once at basal (0), in two equal splits at basal and 30 days after sowing and in three equal splits at 0, 30 and 60 days after sowing. Higher values of ammonium-N (7.2–10%) at all stages of sampling in neem-leaf treated soil than their corresponding control proved neem leaf as an effective nitrification inhibitor. Barring few exceptions, transformation/decrease in amino acid-, hexosamine-, total hydrolysable- and total-N was governed by the rate and frequency of neem-leaf dust application. Irrespective of dose and frequency of neem-leaf application there was a net build up of total non-hydrolysable-N. Both dose and frequency of neem-leaf dust application independently had a significant positive impact on dry matter yield and nitrogen uptake by maize. Application of neem-leaf dust at the rate of 600 mg kg−1 of soil in two equal splits was found to be the best combination for dry matter production as well as nitrogen nutrition of maize. © 2020, NAAS (National Academy of Agricultural Sciences).

Agricultural Research

Effect of Dose and Residence Time of Neem-Leaf Dust as Nitrification Inhibitor on N-Transformation in Maize Soil

While deep neural network-based models have demonstrated compelling performance on various tasks in computer vision and other fields, they have been found to be vulnerable to adversarial attacks. Particularly, deep convolutional neural network (CNN)-based models can be easily fooled by adding a small quasi-imperceptible perturbation to the input, thus resulting in significant drop in prediction accuracies. While most of the previous works have focused on generating one adversary/perturbation per model, it was recently shown that it is possible to learn a continuous distribution over adversarial perturbations for a model. Building upon this work, in this paper, we propose a new technique for image-specific adversary selection and treat it as a retrieval task. The proposed technique utilizes a learned model that ranks the perturbations in a given set of perturbations based on their ability to fool with respect to a given sample. This model is a conditional determinantal point process model that also explicitly induces diversity among the retrieved perturbations. We conduct experiments on the ImageNet dataset using four popular deep CNN image classification models, and demonstrate that the proposed method consistently achieves state-of-the-art fooling rates. © 2019, Springer-Verlag London Ltd., part of Springer Nature.

International Journal of Multimedia Information Retrieval

A retrieval-based approach for diverse and image-specific adversary selection

Optical Flow is a popular method of computer vision for motion estimation. In this paper, we present a refined optical flow estimation method. Central to our approach is exploiting contour information as most of the motion lies on the edges. Further, we have formulated it as sparse to dense motion estimation. Proposed method has been evaluated on challenging real life image sequences of KITTI and Fish4Knowledge database. Results demonstrate that method performs well in case of low contrast, highly cluttered background, dynamic background, occlusion and illumination change. © 2020, Springer Nature Singapore Pte Ltd.