Search Results for author: Omesh Tickoo

Found 18 papers, 3 papers with code

Adaptive Convolutional Neural Networks

1 code implementation ICLR 2019 Julio Cesar Zamora, Jesus Adan Cruz Vargas, Omesh Tickoo

The quest for increased visual recognition performance has led to the development of highly complex neural networks with very deep topologies.

Mitigating Sampling Bias and Improving Robustness in Active Learning

no code implementations13 Sep 2021 Ranganath Krishnan, Alok Sinha, Nilesh Ahuja, Mahesh Subedar, Omesh Tickoo, Ravi Iyer

This paper presents simple and efficient methods to mitigate sampling bias in active learning while achieving state-of-the-art accuracy and model robustness.

Active Learning

Improving Robustness and Efficiency in Active Learning with Contrastive Loss

no code implementations13 Sep 2021 Ranganath Krishnan, Nilesh Ahuja, Alok Sinha, Mahesh Subedar, Omesh Tickoo, Ravi Iyer

This paper introduces supervised contrastive active learning (SCAL) by leveraging the contrastive loss for active learning in a supervised setting.

Active Learning Out-of-Distribution Detection

Partially-supervised novel object captioning leveraging context from paired data

no code implementations10 Sep 2021 Shashank Bujimalla, Mahesh Subedar, Omesh Tickoo

In this paper, we propose an approach to improve image captioning solutions for images with novel objects that do not have caption labels in the training dataset.

Image Captioning Object Detection

Data augmentation to improve robustness of image captioning solutions

no code implementations10 Jun 2021 Shashank Bujimalla, Mahesh Subedar, Omesh Tickoo

In this paper, we study the impact of motion blur, a common quality flaw in real world images, on a state-of-the-art two-stage image captioning solution, and notice a degradation in solution performance as blur intensity increases.

Data Augmentation Image Captioning +1

Energy-Based Anomaly Detection and Localization

no code implementations7 May 2021 Ergin Utku Genc, Nilesh Ahuja, Ibrahima J Ndiour, Omesh Tickoo

This brief sketches initial progress towards a unified energy-based solution for the semi-supervised visual anomaly detection and localization problem.

Anomaly Detection

Improving model calibration with accuracy versus uncertainty optimization

1 code implementation NeurIPS 2020 Ranganath Krishnan, Omesh Tickoo

Obtaining reliable and accurate quantification of uncertainty estimates from deep neural networks is important in safety-critical applications.

Image Classification Variational Inference

Out-Of-Distribution Detection With Subspace Techniques And Probabilistic Modeling Of Features

no code implementations8 Dec 2020 Ibrahima Ndiour, Nilesh Ahuja, Omesh Tickoo

We also show that the feature reconstruction error, which is the $L_2$-norm of the difference between the original feature and the pre-image of its embedding, is highly effective for OOD detection and in some cases superior to the log-likelihood scores.

Dimensionality Reduction Out-of-Distribution Detection

B-SCST: Bayesian Self-Critical Sequence Training for Image Captioning

no code implementations6 Apr 2020 Shashank Bujimalla, Mahesh Subedar, Omesh Tickoo

The "baseline" for the policy-gradients in B-SCST is generated by averaging predictive quality metrics (CIDEr-D) of the captions drawn from the distribution obtained using a Bayesian DNN model.

Bayesian Inference Image Captioning +1

Tree pyramidal adaptive importance sampling

no code implementations18 Dec 2019 Javier Felip, Nilesh Ahuja, Omesh Tickoo

After each new sample operation, a set of tree leaves are subdivided for improving the approximation of the proposal distribution to the target density.

Deep Probabilistic Models to Detect Data Poisoning Attacks

no code implementations3 Dec 2019 Mahesh Subedar, Nilesh Ahuja, Ranganath Krishnan, Ibrahima J. Ndiour, Omesh Tickoo

In the second approach, we use Bayesian deep neural networks trained with mean-field variational inference to estimate model uncertainty associated with the predictions.

Data Poisoning Variational Inference

Probabilistic Modeling of Deep Features for Out-of-Distribution and Adversarial Detection

no code implementations25 Sep 2019 Nilesh A. Ahuja, Ibrahima Ndiour, Trushant Kalyanpur, Omesh Tickoo

We present a principled approach for detecting out-of-distribution (OOD) and adversarial samples in deep neural networks.

Adversarial Attack

Specifying Weight Priors in Bayesian Deep Neural Networks with Empirical Bayes

2 code implementations12 Jun 2019 Ranganath Krishnan, Mahesh Subedar, Omesh Tickoo

We propose MOdel Priors with Empirical Bayes using DNN (MOPED) method to choose informed weight priors in Bayesian neural networks.

Activity Recognition Audio Classification +4

Real-time Approximate Bayesian Computation for Scene Understanding

no code implementations22 May 2019 Javier Felip, Nilesh Ahuja, David Gómez-Gutiérrez, Omesh Tickoo, Vikash Mansinghka

The underlying generative models are built from realistic simulation software, wrapped in a Bayesian error model for the gap between simulation outputs and real data.

Scene Understanding

Uncertainty aware audiovisual activity recognition using deep Bayesian variational inference

no code implementations27 Nov 2018 Mahesh Subedar, Ranganath Krishnan, Paulo Lopez Meyer, Omesh Tickoo, Jonathan Huang

In the multimodal setting, the proposed framework improved precision-recall AUC by 10. 2% on the subset of MiT dataset as compared to non-Bayesian baseline.

Bayesian Inference Multimodal Activity Recognition +1

BAR: Bayesian Activity Recognition using variational inference

no code implementations8 Nov 2018 Ranganath Krishnan, Mahesh Subedar, Omesh Tickoo

We show that the Bayesian inference applied to DNNs provide reliable confidence measures for visual activity recognition task as compared to conventional DNNs.

Activity Recognition Bayesian Inference +1

A Greedy Part Assignment Algorithm for Real-time Multi-person 2D Pose Estimation

no code implementations30 Aug 2017 Srenivas Varadarajan, Parual Datta, Omesh Tickoo

We propose a greedy part assignment algorithm that exploits the inherent structure of the human body to achieve a lower complexity, compared to any of the prior published works.

Pose Estimation

Cannot find the paper you are looking for? You can Submit a new open access paper.