We propose a transfer learning-based solution for the problem of multiple class novelty detection. In particular, we propose an end-to-end deep-learning based approach in which we investigate how the knowledge contained in an external, out-of-distribution dataset can be used to improve the performance of a deep network for visual novelty detection. Our solution differs from the standard deep classification networks on two accounts. First, we use a novel loss function, membership loss, in addition to the classical cross-entropy loss for training networks. Secondly, we use the knowledge from the external dataset more effectively to learn globally negative filters, filters that respond to generic objects outside the known class set. Conference Version (CVPR '19 )

We present a novel model called OCGAN for the classical problem of one-class novelty detection, where, given a set of examples from a particular class, the goal is to determine if a query example is from the same class. Our solution is based on learning latent representations of in-class examples using a denoising auto-encoder network. The key contribution of our work is our proposal to explicitly constrain the latent space to exclusively represent the given class. Conference Version (CVPR '19 )

With the advent of mobile devices a mobile device may be accessed by more than a single enrolled user. In this context, verification of multiple enrolled users has a practical importance. We address the issue of performance degradation associated with multiple user authentication as compared to single user authentication. Journal Version (TIFS '18 ). Conference Version (FG '17 )

Due to unavailability of training samples from negative classes, AA can be viewed as a one-class classification problem. In this work we introduce a Single-class Minimax Probability Machine(1-MPM) based solution called Dual Minimax Probability Machines(DMPM) for AA applications. In this work, we learn an additional hyper-plane to separate training data from the origin by taking into account maximum data covariance. Further, we consider the possibility of modeling the underline distribution of training data as a collection of sub-distributions. Conference Version (BTAS '18 )

In unsupervised image-to-image translation, the goal is to learn the mapping between an input image and an output image using a set of unpaired training images. In this paper, we propose an extension of the unsupervised image-to-image translation problem to multiple input setting. In the paper, we introduce a GAN based framework along witha multi-modal generator structure and a new loss term, latent consistency loss. Conference Version (ICPR '18 Best Student Paper )

We propose a deep learning-based solution for the problem of feature learning in one-class classification. The proposed method operates on top of a Convolutional Neural Network (CNN) of choice and produces descriptive features while maintaining a low intra-class variance in the feature space for the given class. We achieve significant improvements over the state-of-the-art in anomaly detection, novelty detection and mobile active authentication tasks. PDF (arXiv). Code (Github).

We address the problem of quickly detecting intrusions with lower false detection rates in mobile AA systems with higher resource efficiency. Bayesian and Minimax versions of the Quickest Change Detection (QCD) algorithms are introduced to quickly detect intrusions in mobile AA systems. These algorithms are extended with an update rule to facilitate low frequency sensing which leads to low utilization of resources. Journal Version ( TIFS '17). Conference Version (BTAS '16 ). In Media (CBS news).

In this work, we propose to improve the performance of mobile Active Authentication (AA) systems in the low false alarm region using the statistical Extreme Value Theory (EVT). The problem is studied under a Bayesian framework where extremal observations that contribute to mis-verification are given more prominence. We propose modeling the tail of the match distribution using a Generalized Pareto Distribution (GPD) in order to make better inferences about the extremal observations. A method based on the mean excess function is introduced for parameter estimation of the GPD. Conference Version (FG '17 )