arXiv subjects

Gaurav Sharma

Publications and source records attributed to Gaurav Sharma.

64 records · Page 4Linked to original sources

arXiv2016

Latent Embeddings for Zero-shot Classification

We present a novel latent embedding model for learning a compatibility function between image and class embeddings, in the context of zero-shot classification. The proposed method augments the state-of-the-art bilinear compatibility model by incorporating latent variables. Instead of learning a single bilinear map, it learns a collection of maps with the selection, of which map to use, being a latent variable for the current image-class pair. We train the model with a ranking based objective function which penalizes incorrect rankings of the true class for a given image. We empirically demonstrate that our model improves the state-of-the-art for various class embeddings consistently on three challenging publicly available datasets for the zero-shot setting. Moreover, our method leads to visually highly interpretable results with clear clusters of different fine-grained object properties that correspond to different latent variable maps.

cs.CV↗

arXiv2016

LOMo: Latent Ordinal Model for Facial Analysis in Videos

We study the problem of facial analysis in videos. We propose a novel weakly supervised learning method that models the video event (expression, pain etc.) as a sequence of automatically mined, discriminative sub-events (eg. onset and offset phase for smile, brow lower and cheek raise for pain). The proposed model is inspired by the recent works on Multiple Instance Learning and latent SVM/HCRF- it extends such frameworks to model the ordinal or temporal aspect in the videos, approximately. We obtain consistent improvements over relevant competitive baselines on four challenging and publicly available video based facial analysis datasets for prediction of expression, clinical pain and intent in dyadic conversations. In combination with complimentary features, we report state-of-the-art results on these datasets.

cs.CV↗

arXiv2016

Expanded Parts Model for Semantic Description of Humans in Still Images

We introduce an Expanded Parts Model (EPM) for recognizing human attributes (e.g. young, short hair, wearing suit) and actions (e.g. running, jumping) in still images. An EPM is a collection of part templates which are learnt discriminatively to explain specific scale-space regions in the images (in human centric coordinates). This is in contrast to current models which consist of a relatively few (i.e. a mixture of) 'average' templates. EPM uses only a subset of the parts to score an image and scores the image sparsely in space, i.e. it ignores redundant and random background in an image. To learn our model, we propose an algorithm which automatically mines parts and learns corresponding discriminative templates together with their respective locations from a large number of candidate parts. We validate our method on three recent challenging datasets of human attributes and actions. We obtain convincing qualitative and state-of-the-art quantitative results on the three datasets.

cs.CV↗

arXiv2016

X-ray spectroscopy technique for the pile-up region

We report a pile-up rejection technique based on X-ray absorption concept of Beer-Lambert law for measuring true events in the pile-up region. We have detected a 10^4 times weaker peak in the pile-up region. This technique also enables one to resolve the weak peaks adjacent to an intense peak provided the later lies in the lower energy side, and the peaks are at least theoretically resolvable by the detector used. We have resolved such peaks by reducing the intensity ratios in our experiment. The technique allows us to obtain the actual intensities of the observed peaks to have been measured without any attenuator. The possible applications of this technique can be to study the physics of two electron one-photon transition as well as the properties of projectile-like or target-like ions

physics.ins-det↗

arXiv2015

Surface wake field model of beam-foil circular Rydberg states

Production of projectile Rydberg states in fast ion-solid collisions in H-like ions exhibits a pronounce target thickness dependence in spite of these states forming at the last layers. This occurs due to important role of the surface wake field which varies with the target foil thickness. Further, according to the proposed model Rydberg states with low angular momentum are transformed into a circular Rydberg states while passing through the field. The transfer occurs by a single multiphoton process with high probability depending upon the projectile ion velocity with respect to the Fermi velocity of the target electrons.

physics.atom-ph↗

arXiv2015

Local Higher-Order Statistics (LHS) describing images with statistics of local non-binarized pixel patterns

We propose a new image representation for texture categorization and facial analysis, relying on the use of higher-order local differential statistics as features. It has been recently shown that small local pixel pattern distributions can be highly discriminative while being extremely efficient to compute, which is in contrast to the models based on the global structure of images. Motivated by such works, we propose to use higher-order statistics of local non-binarized pixel patterns for the image description. The proposed model does not require either (i) user specified quantization of the space (of pixel patterns) or (ii) any heuristics for discarding low occupancy volumes of the space. We propose to use a data driven soft quantization of the space, with parametric mixture models, combined with higher-order statistics, based on Fisher scores. We demonstrate that this leads to a more expressive representation which, when combined with discriminatively learned classifiers and metrics, achieves state-of-the-art performance on challenging texture and facial analysis datasets, in low complexity setup. Further, it is complementary to higher complexity features and when combined with them improves performance.

cs.CV↗

arXiv2015

Scalable Nonlinear Embeddings for Semantic Category-based Image Retrieval

We propose a novel algorithm for the task of supervised discriminative distance learning by nonlinearly embedding vectors into a low dimensional Euclidean space. We work in the challenging setting where supervision is with constraints on similar and dissimilar pairs while training. The proposed method is derived by an approximate kernelization of a linear Mahalanobis-like distance metric learning algorithm and can also be seen as a kernel neural network. The number of model parameters and test time evaluation complexity of the proposed method are O(dD) where D is the dimensionality of the input features and d is the dimension of the projection space - this is in contrast to the usual kernelization methods as, unlike them, the complexity does not scale linearly with the number of training examples. We propose a stochastic gradient based learning algorithm which makes the method scalable (w.r.t. the number of training examples), while being nonlinear. We train the method with up to half a million training pairs of 4096 dimensional CNN features. We give empirical comparisons with relevant baselines on seven challenging datasets for the task of low dimensional semantic category based image retrieval.

cs.CV↗

arXiv2012

Reliable Resource Selection in Grid Environment

The primary concern in area of computational grid is security and resources. Most of the existing grids address this problem by authenticating the users, hosts and their interactions in an appropriate manner. A secured system is compulsory for the efficient utilization of grid services. The high degree of strangeness has been identified as the problem factors in the secured selection of grid. Without the assurance of a higher degree of trust relationship, competent resource selection and utilization cannot be achieved. In this paper we proposed an approach which is providing reliability and reputation aware security for resource selection in grid environment. In this approach, the self-protection capability and reputation weightage is utilized to obtain the Reliability Factor (RF) value. Therefore jobs are allocated to the resources that posses higher RF values. Extensive experimental evaluation shows that as higher trust and reliable nodes are selected the chances of failure decreased drastically.

cs.DC↗

arXiv2007

Prototyping Bio-Nanorobots using Molecular Dynamics Simulation

This paper presents a molecular mechanics study using a molecular dynamics software (NAMD) coupled to virtual reality (VR) techniques for intuitive Bio-NanoRobotic prototyping. Using simulated Bio-Nano environments in VR, the operator can design and characterize through physical simulation and 3-D visualization the behavior of Bio-NanoRobotic components and structures. The main novelty of the proposed simulations is based on the characterization of stiffness performances of passive joints-based deca-alanine protein molecule and active joints-based viral protein motor (VPL) in their native environment. Their use as elementary Bio-NanoRobotic components (1 dof platform) are also simulated and the results discussed.

physics.bio-ph↗

arXiv2001

Digital Color Imaging

This paper surveys current technology and research in the area of digital color imaging. In order to establish the background and lay down terminology, fundamental concepts of color perception and measurement are first presented us-ing vector-space notation and terminology. Present-day color recording and reproduction systems are reviewed along with the common mathematical models used for representing these devices. Algorithms for processing color images for display and communication are surveyed, and a forecast of research trends is attempted. An extensive bibliography is provided.

cs.CV↗