Search Results for author: Prakash Ishwar

Found 35 papers, 15 papers with code

Detecting Frames in News Headlines and Lead Images in U.S. Gun Violence Coverage

no code implementations Findings (EMNLP) 2021 Isidora Tourni, Lei Guo, Taufiq Husada Daryanto, Fabian Zhafransyah, Edward Edberg Halim, Mona Jalal, Boqi Chen, Sha Lai, Hengchang Hu, Margrit Betke, Prakash Ishwar, Derry Tanti Wijaya

Such perspectives are called “frames” in communication research. We study, for the first time, the value of combining lead images and their contextual information with text to identify the frame of a given news article.

Multimodal Text and Image Classification News Annotation +1

An Unsupervised Approach to Discover Media Frames

1 code implementation PoliticalNLP (LREC) 2022 Sha Lai, Yanru Jiang, Lei Guo, Margrit Betke, Prakash Ishwar, Derry Tanti Wijaya

We discuss the effectiveness of our approach by comparing the frames it generates in an unsupervised manner to the domain-expert-derived frames for the issue of gun violence, for which a supervised learning model for frame recognition exists.

Community Detection

On neural and dimensional collapse in supervised and unsupervised contrastive learning with hard negative sampling

no code implementations9 Nov 2023 Ruijie Jiang, Thuan Nguyen, Shuchin Aeron, Prakash Ishwar

For a widely-studied data model and general loss and sample-hardening functions we prove that the Supervised Contrastive Learning (SCL), Hard-SCL (HSCL), and Unsupervised Contrastive Learning (UCL) risks are minimized by representations that exhibit Neural Collapse (NC), i. e., the class means form an Equianglular Tight Frame (ETF) and data from the same class are mapped to the same representation.

Contrastive Learning

A principled approach to model validation in domain generalization

1 code implementation2 Apr 2023 Boyang Lyu, Thuan Nguyen, Matthias Scheutz, Prakash Ishwar, Shuchin Aeron

Domain generalization aims to learn a model with good generalization ability, that is, the learned model should not only perform well on several seen domains but also on unseen domains with different data distributions.

Classification Domain Generalization +1

Spatio-Visual Fusion-Based Person Re-Identification for Overhead Fisheye Images

no code implementations22 Dec 2022 Mertcan Cokbas, Prakash Ishwar, Janusz Konrad

Person re-identification (PRID) has been thoroughly researched in typical surveillance scenarios where various scenes are monitored by side-mounted, rectilinear-lens cameras.

Person Re-Identification

Trade-off between reconstruction loss and feature alignment for domain generalization

1 code implementation26 Oct 2022 Thuan Nguyen, Boyang Lyu, Prakash Ishwar, Matthias Scheutz, Shuchin Aeron

To deal with challenging settings in DG where both data and label of the unseen domain are not available at training time, the most common approach is to design the classifiers based on the domain-invariant representation features, i. e., the latent representations that are unchanged and transferable between domains.

Domain Generalization Transfer Learning

FRIDA: Fisheye Re-Identification Dataset with Annotations

no code implementations4 Oct 2022 Mertcan Cokbas, John Bolognino, Janusz Konrad, Prakash Ishwar

Person re-identification (PRID) from side-mounted rectilinear-lens cameras is a well-studied problem.

Person Re-Identification

Supervised Contrastive Learning with Hard Negative Samples

1 code implementation31 Aug 2022 Ruijie Jiang, Thuan Nguyen, Prakash Ishwar, Shuchin Aeron

In this paper, motivated by the effectiveness of hard-negative sampling strategies in H-UCL and the usefulness of label information in SCL, we propose a contrastive learning framework called hard-negative supervised contrastive learning (H-SCL).

Contrastive Learning Self-Supervised Learning

Joint covariate-alignment and concept-alignment: a framework for domain generalization

1 code implementation1 Aug 2022 Thuan Nguyen, Boyang Lyu, Prakash Ishwar, Matthias Scheutz, Shuchin Aeron

Particularly, our framework proposes to jointly minimize both the covariate-shift as well as the concept-shift between the seen domains for a better performance on the unseen domain.

Concept Alignment Domain Generalization

Conditional entropy minimization principle for learning domain invariant representation features

2 code implementations25 Jan 2022 Thuan Nguyen, Boyang Lyu, Prakash Ishwar, Matthias Scheutz, Shuchin Aeron

Invariance-principle-based methods such as Invariant Risk Minimization (IRM), have recently emerged as promising approaches for Domain Generalization (DG).

Domain Generalization

Hard Negative Sampling via Regularized Optimal Transport for Contrastive Representation Learning

1 code implementation4 Nov 2021 Ruijie Jiang, Prakash Ishwar, Shuchin Aeron

We analyze a novel min-max framework that seeks a representation which minimizes the maximum (worst-case) generalized contrastive learning loss over all couplings (joint distributions between positive and negative samples subject to marginal constraints) and prove that the resulting min-max optimum representation will be degenerate.

Contrastive Learning Representation Learning

Ergodic Limits, Relaxations, and Geometric Properties of Random Walk Node Embeddings

no code implementations9 Sep 2021 Christy Lin, Daniel Sussman, Prakash Ishwar

This paper studies properties of random walk based node embeddings in the unsupervised setting of discovering hidden block structure in the network, i. e., learning node representations whose cluster structure in Euclidean space reflects their adjacency structure within the network.

Link Prediction Node Classification +1

Barycentric-alignment and reconstruction loss minimization for domain generalization

1 code implementation4 Sep 2021 Boyang Lyu, Thuan Nguyen, Prakash Ishwar, Matthias Scheutz, Shuchin Aeron

To bridge this gap between theory and practice, we introduce a new upper bound that is free of terms having such dual dependence, resulting in a fully optimizable risk upper bound for the unseen domain.

Domain Generalization Representation Learning

BSUV-Net 2.0: Spatio-Temporal Data Augmentations for Video-Agnostic Supervised Background Subtraction

1 code implementation23 Jan 2021 M. Ozan Tezcan, Prakash Ishwar, Janusz Konrad

In this work, we introduce spatio-temporal data augmentations and apply them to one of the leading video-agnostic BGS algorithms, BSUV-Net.

OpenFraming: We brought the ML; you bring the data. Interact with your data and discover its frames

2 code implementations16 Aug 2020 Alyssa Smith, David Assefa Tofu, Mona Jalal, Edward Edberg Halim, Yimeng Sun, Vidya Akavoor, Margrit Betke, Prakash Ishwar, Lei Guo, Derry Wijaya

The degree of user involvement is flexible: they can run models that have been pre-trained on select issues; submit labeled documents and train a new model for frame classification; or submit unlabeled documents and obtain potential frames of the documents.

General Classification

Multi-Label and Multilingual News Framing Analysis

no code implementations ACL 2020 Afra Feyza Aky{\"u}rek, Lei Guo, R Elanwar, a, Prakash Ishwar, Margrit Betke, Derry Tanti Wijaya

News framing refers to the practice in which aspects of specific issues are highlighted in the news to promote a particular interpretation.

Transfer Learning Translation

RAPiD: Rotation-Aware People Detection in Overhead Fisheye Images

1 code implementation23 May 2020 Zhihao Duan, M. Ozan Tezcan, Hayato Nakamura, Prakash Ishwar, Janusz Konrad

Recent methods for people detection in overhead, fisheye images either use radially-aligned bounding boxes to represent people, assuming people always appear along image radius or require significant pre-/post-processing which radically increases computational complexity.

Low-Resolution Overhead Thermal Tripwire for Occupancy Estimation

no code implementations12 Apr 2020 Mertcan Cokbas, Prakash Ishwar, Janusz Konrad

We propose a people counting system which uses a low-resolution thermal sensor.

VAE/WGAN-Based Image Representation Learning For Pose-Preserving Seamless Identity Replacement In Facial Images

no code implementations2 Mar 2020 Hiroki Kawai, Jia-Wei Chen, Prakash Ishwar, Janusz Konrad

We present a novel variational generative adversarial network (VGAN) based on Wasserstein loss to learn a latent representation from a face image that is invariant to identity but preserves head-pose information.

Representation Learning

BSUV-Net: A Fully-Convolutional Neural Network forBackground Subtraction of Unseen Videos

1 code implementation ICCV 2020 M. Ozan Tezcan, Prakash Ishwar, Janusz Konrad

In order to reduce the chance of overfitting, we also introduce a new data-augmentation technique which mitigates the impact of illumination difference between the background frames and the current frame.

Data Augmentation Object Tracking

BSUV-Net: A Fully-Convolutional Neural Network for Background Subtraction of Unseen Videos

1 code implementation26 Jul 2019 M. Ozan Tezcan, Prakash Ishwar, Janusz Konrad

In order to reduce the chance of overfitting, we also introduce a new data-augmentation technique which mitigates the impact of illumination difference between the background frames and the current frame.

Data Augmentation Object Tracking +2

A Cyclically-Trained Adversarial Network for Invariant Representation Learning

no code implementations21 Jun 2019 Jiawei Chen, Janusz Konrad, Prakash Ishwar

Specifically, we propose a cyclically-trained adversarial network to learn a mapping from image space to latent representation space and back such that the latent representation is invariant to a specified factor of variation (e. g., identity).

Representation Learning

BUOCA: Budget-Optimized Crowd Worker Allocation

no code implementations11 Jan 2019 Mehrnoosh Sameki, Sha Lai, Kate K. Mays, Lei Guo, Prakash Ishwar, Margrit Betke

We next train a machine learning system (BUOCA-ML) that predicts an optimal number of crowd workers needed to maximize the accuracy of the labeling.

Privacy-Preserving Adversarial Networks

no code implementations19 Dec 2017 Ardhendu Tripathy, Ye Wang, Prakash Ishwar

We propose a data-driven framework for optimizing privacy-preserving data release mechanisms to attain the information-theoretically optimal tradeoff between minimizing distortion of useful data and concealing specific sensitive information.

Privacy Preserving

Node Embedding via Word Embedding for Network Community Discovery

1 code implementation9 Nov 2016 Weicong Ding, Christy Lin, Prakash Ishwar

Neural node embeddings have recently emerged as a powerful representation for supervised learning tasks involving graph-structured data.

Clustering Graph Generation

Semi-Coupled Two-Stream Fusion ConvNets for Action Recognition at Extremely Low Resolutions

no code implementations12 Oct 2016 Jiawei Chen, Jonathan Wu, Janusz Konrad, Prakash Ishwar

Deep convolutional neural networks (ConvNets) have been recently shown to attain state-of-the-art performance for action recognition on standard-resolution videos.

Action Recognition Temporal Action Localization

Necessary and Sufficient Conditions and a Provably Efficient Algorithm for Separable Topic Discovery

no code implementations23 Aug 2015 Weicong Ding, Prakash Ishwar, Venkatesh Saligrama

We develop necessary and sufficient conditions and a novel provably consistent and efficient algorithm for discovering topics (latent factors) from observations (documents) that are realized from a probabilistic mixture of shared latent factors that have certain properties.

Topic Models

Learning Mixed Membership Mallows Models from Pairwise Comparisons

no code implementations3 Apr 2015 Weicong Ding, Prakash Ishwar, Venkatesh Saligrama

Our key algorithmic insight for estimation is to establish a statistical connection between M4 and topic models by viewing pairwise comparisons as words, and users as documents.

Topic Models

A Topic Modeling Approach to Ranking

no code implementations11 Dec 2014 Weicong Ding, Prakash Ishwar, Venkatesh Saligrama

We propose a topic modeling approach to the prediction of preferences in pairwise comparisons.

Sensing-Aware Kernel SVM

no code implementations2 Dec 2013 Weicong Ding, Prakash Ishwar, Venkatesh Saligrama, W. Clem Karl

We propose a novel approach for designing kernels for support vector machines (SVMs) when the class label is linked to the observation through a latent state and the likelihood function of the observation given the state (the sensing model) is available.

General Classification Image Classification

Necessary and Sufficient Conditions for Novel Word Detection in Separable Topic Models

no code implementations30 Oct 2013 Weicong Ding, Prakash Ishwar, Mohammad H. Rohban, Venkatesh Saligrama

The simplicial condition and other stronger conditions that imply it have recently played a central role in developing polynomial time algorithms with provable asymptotic consistency and sample complexity guarantees for topic estimation in separable topic models.

Topic Models

An Impossibility Result for High Dimensional Supervised Learning

no code implementations29 Jan 2013 Mohammad Hossein Rohban, Prakash Ishwar, Birant Orten, William C. Karl, Venkatesh Saligrama

We study high-dimensional asymptotic performance limits of binary supervised classification problems where the class conditional densities are Gaussian with unknown means and covariances and the number of signal dimensions scales faster than the number of labeled training samples.

General Classification Vocal Bursts Intensity Prediction

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