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

Rapid advancements in high-throughput technologies have resulted in genome-scale time series datasets. Uncovering the temporal sequence of gene regulatory events, in the form of time-varying gene regulatory networks (GRNs), demands computationally fast, accurate, and scalable algorithms. The existing algorithms can be divided into two categories: ones that are time-intensive and hence unscalable; and others that impose structural constraints to become scalable. In this paper, a novel algorithm, namely 'an algorithm for reconstructing Time-varying Gene regulatory networks with Shortlisted candidate regulators' (TGS), is proposed. TGS is time-efficient and does not impose any structural constraints. Moreover, it provides such flexibility and time-efficiency, without losing its accuracy. TGS consistently outperforms the state-of-the-art algorithms in true positive detection, on three benchmark synthetic datasets. However, TGS does not perform as well in false positive rejection. To mitigate this issue, TGS+ is proposed. TGS+ demonstrates competitive false positive rejection power, while maintaining the superior speed and true positive detection power of TGS. Nevertheless, the main memory requirements of both TGS variants grow exponentially with the number of genes, which they tackle by restricting the maximum number of regulators for each gene. Relaxing this restriction remains a challenge as the actual number of regulators is not known a priori. © 2004-2012 IEEE.

World University root node

Fulltext

IEEE/ACM Transactions on Computational Biology and Bioinformatics

Rapid Reconstruction of Time-Varying Gene Regulatory Networks

High resolution Digital Elevation Models(DEMs) are an important requirement for many applications like modelling water flow, landslides, avalanches etc. Yet publicly available DEMs have low resolution for most parts of the world. Despite tremendous success in image super-resolution task using deep learning solutions, there are very few works that have used these powerful systems on DEMs to generate HRDEMs. Motivated from feedback neural networks, we propose a novel neural network architecture that learns to add high frequency details iteratively to low resolution DEM, turning it into a high resolution DEM without compromising its fidelity. Our experiments confirm that without any additional modality such as aerial images(RGB), our network DSRFB achieves RMSEs of 0.59 to 1.27 across 4 different terrains having diverse geographical structures. © 2020 IEEE.

International Geoscience and Remote Sensing Symposium (IGARSS)

Feedback Neural Network Based Super-Resolution of DEM for Generating High Fidelity Features

Passive Batch Injection Training Technique: Boosting Network Performance by Injecting Mini-Batches from a different Data Distribution

Cpwc: Contextual point wise convolution for object recognition

A graph is called a split graph if its vertex set can be partitioned into a clique and an independent set. Split graphs have rich mathematical structure and interesting algorithmic properties making it one of the most well-studied special graph classes. In the SPLIT VERTEX DELETION problem, given a graph and a positive integer k, the objective is to test whether there exists a subset of at most k vertices whose deletion results in a split graph. In this paper, we design a kernel for this problem with O(k2) vertices, improving upon the previous cubic bound known. Also, by giving a simple reduction from the VERTEX COVER problem, we establish that SPLIT VERTEX DELETION does not admit a kernel with O(k2−ϵ) edges, for any ϵ&gt;0, unless NP⊆coNP/poly. © 2020 Elsevier B.V.

Theoretical Computer Science

Quadratic vertex kernel for split vertex deletion

In a multiwinner election based on the Condorcet criterion, we are given a set of candidates, and a set of voters with strict preference rankings over the candidates. A committee is weakly Gehrlein stable (WGS) if each committee member is preferred to each non-member by at least half of the voters. Recently, Aziz et al. [IJCAI 2017] studied the computational complexity of finding a WGS committee of size k. They show that this problem is NP-hard in general and polynomial-time solvable when the number of voters is odd. In this article, we initiate a systematic study of the problem in the realm of parameterized complexity. We first show that the problem is W[1]-hard when parameterized by the size of the committee. To overcome this intractability result, we use a known reformulation of WGS as a problem on directed graphs and then use parameters that measure the “structure” of these directed graphs. We show that the problem is fixed parameter tractable and admits linear kernels with respect to these parameters; and also present an exact-exponential time algorithm with running in time O(1. 2207 nnO ( 1 )) , where n denotes the number of candidates. © 2020, Springer Science+Business Media, LLC, part of Springer Nature.

Autonomous Agents and Multi-Agent Systems

Gehrlein stability in committee selection: parameterized hardness and algorithms

This unique volume reviews the latest advances in domain adaptation in the training of machine learning algorithms for visual understanding, offering valuable insights from an international selection of experts in the field. The text presents a diverse selection of novel techniques, covering applications of object recognition, face recognition, and action and event recognition. © Springer Nature Switzerland AG 2020. All rights reserved.

Domain Adaptation for Visual Understanding

Domain adaptation for visual understanding

“By reading only the title and abstract, do you think this research will be accepted in an AI conference?” A common impromptu reply would be “I don’t know but I have an intuition that this research might get accepted”. Intuition is often employed by humans to solve challenging problems without explicit efforts. Intuition is not trained but is learned from one’s own experience and observation. The aim of this research is to provide intuition to an algorithm, apart from what they are trained to know in a supervised manner. We present a novel intuition learning framework that learns to perform a task completely from unlabeled data. The proposed framework uses a continuous state reinforcement learning mechanism to learn a feature representation and a data-label mapping function using unlabeled data. The mapping functions and feature representation are succinct and can be used to supplement any supervised or semi-supervised algorithm. The experiments on the CIFAR-10 database show challenging cases where intuition learning improves the performance of a given classifier. © Springer Nature Switzerland AG 2020. All rights reserved.

Intuition learning

With the availability of smartphone cameras, high speed internet, and connectivity to social media, users post content on the go including check-ins, text, and images. Privacy leaks due to posts related to check-ins and text is an issue in itself, however, this paper discusses the potential leak of one's biometric information via images posted on social media. While posting photos of themselves or highlighting miniature objects, users end up posting content that leads to an irreversible loss of biometric information such as ocular region, fingerprint, knuckle print, and ear print. In this paper, we discuss the effect of the loss of the finger-selfie details from social media. We demonstrate that this could potentially lead to matching finger-selfies with livescan fingerprints. Further, to prevent the leak of the finger-selfie details, we propose privacy preserving adversarial learning algorithm. The algorithm learns a perturbation to prevent the misuse of finger-selfie towards recognition, yet keeping the visual quality intact to highlight the minuscule object. The experiments are presented on the ISPFDv1 database. Further, we propose a new publicly available Social-Media Posted Finger-selfie (SMPF) Database, containing 1, 000 finger-selfie images posted on Instagram. © 2020 IEEE.

IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops

On privacy preserving anonymization of finger-selfies

The widespread applicability of deep learning based algorithms demands dedicated attention towards ensuring unbiased behavior. Biased feature learning (for or against a particular sub-group) might often result in unfair predictions. In order to address the above issue, this research proposes a novel Filter-Drop algorithm for learning unbiased representations. The proposed technique focuses on learning the features useful for predicting the biasing attribute (or the sensitive attribute), followed by their elimination while performing the primary classification task. To this effect, a multi-task network is trained, which prevents the features capturing the attribute variations from being used for the primary classification task. The efficacy of the proposed Filter-Drop technique is demonstrated on two facial analysis datasets: UTKFace dataset and FairFace dataset. The proposed technique achieves similar performance across different ethnicity groups while training with highly skewed training data as well. © 2020 IEEE.