no code implementations • 18 Mar 2024 • Jurijs Nazarovs, Zhichun Huang, Xingjian Zhen, Sourav Pal, Rudrasis Chakraborty, Vikas Singh
In this work, we introduce a mechanism to learn the distribution of trajectories by parameterizing the transition function $f$ explicitly as an element in a function space.
no code implementations • 12 Mar 2024 • Zhanpeng Zeng, Karthikeyan Sankaralingam, Vikas Singh
A popular strategy is the use of low bit-width integers to approximate the original entries in a matrix.
1 code implementation • 12 Mar 2024 • Zhanpeng Zeng, Michael Davies, Pranav Pulijala, Karthikeyan Sankaralingam, Vikas Singh
While GPU clusters are the de facto choice for training large deep neural network (DNN) models today, several reasons including ease of workflow, security and cost have led to efforts investigating whether CPUs may be viable for inference in routine use in many sectors of the industry.
no code implementations • 10 Mar 2024 • Harshavardhan Adepu, Zhanpeng Zeng, Li Zhang, Vikas Singh
If quantization is interpreted as the addition of noise, our casting of the problem allows invoking an extensive body of known consistent recovery and noise robustness guarantees.
no code implementations • 5 Mar 2024 • Sotirios Panagiotis Chytas, Vishnu Suresh Lokhande, Peiran Li, Vikas Singh
Further, we discuss how this style of formulation offers a unified perspective on at least 5+ distinct problem settings, from self-supervised learning to matching problems in 3D reconstruction.
1 code implementation • 19 Dec 2023 • Vikas Singh
The present filter has two primary steps: The first stage categorizes images as lightly, medium, and heavily corrupted based on an adaptive threshold by comparing the M-ALD of processed pixels with the upper and lower MF of the type-2 fuzzy identifier.
1 code implementation • 21 Jul 2022 • Zhanpeng Zeng, Sourav Pal, Jeffery Kline, Glenn M Fung, Vikas Singh
Transformers have emerged as a preferred model for many tasks in natural langugage processing and vision.
1 code implementation • 20 Jul 2022 • Xingjian Zhen, Zihang Meng, Rudrasis Chakraborty, Vikas Singh
Comparing the functional behavior of neural network models, whether it is a single network over time or two (or more networks) during or post-training, is an essential step in understanding what they are learning (and what they are not), and for identifying strategies for regularization or efficiency improvements.
1 code implementation • CVPR 2022 • Ronak Mehta, Sourav Pal, Vikas Singh, Sathya N. Ravi
For models which require no training (k-NN), simply deleting the closest original sample can be effective.
1 code implementation • CVPR 2022 • Vishnu Suresh Lokhande, Rudrasis Chakraborty, Sathya N. Ravi, Vikas Singh
Pooling multiple neuroimaging datasets across institutions often enables improvements in statistical power when evaluating associations (e. g., between risk factors and disease outcomes) that may otherwise be too weak to detect.
no code implementations • 19 Feb 2022 • Jurijs Nazarovs, Ronak R. Mehta, Vishnu Suresh Lokhande, Vikas Singh
This is directly related to the structure of the computation graph, which can grow linearly as a function of the number of MC samples needed.
no code implementations • 18 Feb 2022 • Jurijs Nazarovs, Rudrasis Chakraborty, Songwong Tasneeyapant, Sathya N. Ravi, Vikas Singh
Panel data involving longitudinal measurements of the same set of participants taken over multiple time points is common in studies to understand childhood development and disease modeling.
no code implementations • CVPR 2022 • Jurijs Nazarovs, Zhichun Huang, Songwong Tasneeyapant, Rudrasis Chakraborty, Vikas Singh
Quantitative descriptions of confidence intervals and uncertainties of the predictions of a model are needed in many applications in vision and machine learning.
no code implementations • 1 Dec 2021 • Zhichun Huang, Rudrasis Chakraborty, Vikas Singh
Generative models which use explicit density modeling (e. g., variational autoencoders, flow-based generative models) involve finding a mapping from a known distribution, e. g. Gaussian, to the unknown input distribution.
1 code implementation • 18 Nov 2021 • Zhanpeng Zeng, Yunyang Xiong, Sathya N. Ravi, Shailesh Acharya, Glenn Fung, Vikas Singh
In this paper, we show that a Bernoulli sampling attention mechanism based on Locality Sensitive Hashing (LSH), decreases the quadratic complexity of such models to linear.
no code implementations • 29 Sep 2021 • Zhichun Huang, Rudrasis Chakraborty, Vikas Singh
Generative models which use explicit density modeling (e. g., variational autoencoders, flow-based generative models) involve finding a mapping from a known distribution, e. g. Gaussian, to the unknown input distribution.
1 code implementation • ICCV 2021 • Zihang Meng, Vikas Singh, Sathya N. Ravi
We study how stochastic differential equation (SDE) based ideas can inspire new modifications to existing algorithms for a set of problems in computer vision.
1 code implementation • NeurIPS 2021 • Zihang Meng, Rudrasis Chakraborty, Vikas Singh
We present an efficient stochastic algorithm (RSG+) for canonical correlation analysis (CCA) using a reparametrization of the projection matrices.
1 code implementation • CVPR 2021 • Zihang Meng, Licheng Yu, Ning Zhang, Tamara Berg, Babak Damavandi, Vikas Singh, Amy Bearman
Learning the grounding of each word is challenging, due to noise in the human-provided traces and the presence of words that cannot be meaningfully visually grounded.
1 code implementation • CVPR 2021 • Xingjian Zhen, Rudrasis Chakraborty, Vikas Singh
One strategy for adversarially training a robust model is to maximize its certified radius -- the neighborhood around a given training sample for which the model's prediction remains unchanged.
1 code implementation • 16 Feb 2021 • Aditya Kumar Akash, Vishnu Suresh Lokhande, Sathya N. Ravi, Vikas Singh
Learning invariant representations is a critical first step in a number of machine learning tasks.
6 code implementations • 7 Feb 2021 • Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, Vikas Singh
The scalability of Nystr\"{o}mformer enables application to longer sequences with thousands of tokens.
Ranked #13 on Semantic Textual Similarity on MRPC (F1 metric)
no code implementations • 1 Jan 2021 • Zhichun Huang, Rudrasis Chakraborty, Xingjian Zhen, Vikas Singh
Flow-based generative models refer to deep generative models with tractable likelihoods, and offer several attractive properties including efficient density estimation and sampling.
no code implementations • 1 Jan 2021 • Zihang Meng, Rudrasis Chakraborty, Vikas Singh
We present an efficient stochastic algorithm (RSG+) for canonical correlation analysis (CCA) derived via a differential geometric perspective of the underlying optimization task.
no code implementations • NeurIPS 2021 • Zihang Meng, Lopamudra Mukherjee, Vikas Singh, Sathya N. Ravi
We propose a framework which makes it feasible to directly train deep neural networks with respect to popular families of task-specific non-decomposable per- formance measures such as AUC, multi-class AUC, F -measure and others, as well as models such as non-negative matrix factorization.
1 code implementation • 18 Dec 2020 • Xingjian Zhen, Rudrasis Chakraborty, Liu Yang, Vikas Singh
Partly due to this gap, there are also no modality transfer/translation models for manifold-valued data whereas numerous such methods based on generative models are available for natural images.
no code implementations • NeurIPS Workshop TDA_and_Beyond 2020 • Eric Bunch, Qian You, Glenn Fung, Vikas Singh
Recently, neural network architectures have been developed to accommodate when the data has the structure of a graph or, more generally, a hypergraph.
no code implementations • 2 Sep 2020 • Vikas Singh, Homanga Bharadhwaj, Nishchal K. Verma
Clustering techniques have been proved highly suc-cessful for Takagi-Sugeno (T-S) fuzzy model identification.
no code implementations • CVPR 2017 • Won Hwa Kim, Mona Jalal, Seongjae Hwang, Sterling C. Johnson, Vikas Singh
The adoption of "human-in-the-loop" paradigms in computer vision and machine learning is leading to various applications where the actual data acquisition (e. g., human supervision) and the underlying inference algorithms are closely interwined.
no code implementations • 10 Aug 2020 • Vikas Singh, Pooja Agrawal, Teena Sharma, Nishchal K. Verma
The performance of the proposed filter is compared with the various state-of-the-art methods in terms of peak signal-to-noise ratio and computation time.
3 code implementations • 30 Apr 2020 • Zihang Meng, Sathya N. Ravi, Vikas Singh
We describe our development and show the use of our solver in a video segmentation task and meta-learning for few-shot learning.
4 code implementations • CVPR 2021 • Yunyang Xiong, Hanxiao Liu, Suyog Gupta, Berkin Akin, Gabriel Bender, Yongzhe Wang, Pieter-Jan Kindermans, Mingxing Tan, Vikas Singh, Bo Chen
By incorporating regular convolutions in the search space and directly optimizing the network architectures for object detection, we obtain a family of object detection models, MobileDets, that achieve state-of-the-art results across mobile accelerators.
1 code implementation • ECCV 2020 • Vishnu Suresh Lokhande, Aditya Kumar Akash, Sathya N. Ravi, Vikas Singh
We provide a detailed technical analysis and present experiments demonstrating that various fairness measures from the literature can be reliably imposed on a number of training tasks in vision in a manner that is interpretable.
no code implementations • 24 Dec 2019 • Vikas Singh, Nishchal K. Verma
This paper presents a new fuzzy k-means algorithm for the clustering of high dimensional data in various subspaces.
no code implementations • 24 Dec 2019 • Vikas Singh, Nishchal K. Verma
In recent years, intelligent condition-based monitor-ing of rotary machinery systems has become a major researchfocus of machine fault diagnosis.
1 code implementation • ICCV 2019 • Xingjian Zhen, Rudrasis Chakraborty, Nicholas Vogt, Barbara B. Bendlin, Vikas Singh
Efforts are underway to study ways via which the power of deep neural networks can be extended to non-standard data types such as structured data (e. g., graphs) or manifold-valued data (e. g., unit vectors or special matrices).
no code implementations • 26 Sep 2019 • Sathya N. Ravi, Abhay Venkatesh, Glenn Moo Fung, Vikas Singh
Data dependent regularization is known to benefit a wide variety of problems in machine learning.
1 code implementation • CVPR 2020 • Vishnu Suresh Lokhande, Songwong Tasneeyapant, Abhay Venkatesh, Sathya N. Ravi, Vikas Singh
Rectified Linear Units (ReLUs) are among the most widely used activation function in a broad variety of tasks in vision.
1 code implementation • ICCV 2019 • Haoliang Sun, Ronak Mehta, Hao H. Zhou, Zhichun Huang, Sterling C. Johnson, Vivek Prabhakaran, Vikas Singh
Motivated by developments in modality transfer in vision, we study the generation of certain types of PET images from MRI data.
no code implementations • 16 May 2019 • Owen Levin, Zihang Meng, Vikas Singh, Xiaojin Zhu
Recently it's been shown that neural networks can use images of human faces to accurately predict Body Mass Index (BMI), a widely used health indicator.
1 code implementation • ICCV 2019 • Yunyang Xiong, Ronak Mehta, Vikas Singh
In the latter case, the optimization is often non-differentiable and also not very amenable to derivative-free optimization methods.
no code implementations • ICCV 2019 • Seong Jae Hwang, Zirui Tao, Won Hwa Kim, Vikas Singh
Such models may work for cross-sectional studies, however, they are not suitable to generate data for longitudinal studies that focus on "progressive" behavior in a sequence of data.
1 code implementation • ECCV 2018 • Zihang Meng, Nagesh Adluru, Hyunwoo J. Kim, Glenn Fung, Vikas Singh
A sizable body of work on relative attributes provides compelling evidence that relating pairs of images along a continuum of strength pertaining to a visual attribute yields significant improvements in a wide variety of tasks in vision.
no code implementations • 10 Jun 2018 • Hao Henry Zhou, Yunyang Xiong, Vikas Singh
We provide simple schemes to build Bayesian Neural Networks (BNNs), block by block, inspired by a recent idea of computation skeletons.
no code implementations • CVPR 2018 • Seong Jae Hwang, Sathya N. Ravi, Zirui Tao, Hyunwoo J. Kim, Maxwell D. Collins, Vikas Singh
Visual relationships provide higher-level information of objects and their relations in an image â this enables a semantic understanding of the scene and helps downstream applications.
no code implementations • CVPR 2018 • Lopamudra Mukherjee, Sathya N. Ravi, Jiming Peng, Vikas Singh
In this paper, we study the quantization problem in the setting where subspaces are orthogonal and show that this problem is intricately related to a specific type of spectral decomposition of the data.
1 code implementation • NeurIPS 2018 • Rudrasis Chakraborty, Chun-Hao Yang, Xingjian Zhen, Monami Banerjee, Derek Archer, David Vaillancourt, Vikas Singh, Baba C. Vemuri
We show how recurrent statistical recurrent network models can be defined in such spaces.
no code implementations • 19 Apr 2018 • Seong Jae Hwang, Ronak Mehta, Hyunwoo J. Kim, Vikas Singh
There has recently been a concerted effort to derive mechanisms in vision and machine learning systems to offer uncertainty estimates of the predictions they make.
4 code implementations • 21 Mar 2018 • Sathya N. Ravi, Ronak Mehta, Vikas Singh
We revisit the Blind Deconvolution problem with a focus on understanding its robustness and convergence properties.
1 code implementation • 17 Mar 2018 • Sathya N. Ravi, Tuan Dinh, Vishnu Sai Rao Lokhande, Vikas Singh
We provide convergence guarantees and show a suite of immediate benefits that are possible -- from training ResNets with fewer layers but better accuracy simply by substituting in our version of CG to faster training of GANs with 50% fewer epochs in image inpainting applications to provably better generalization guarantees using efficiently implementable forms of recently proposed regularizers.
no code implementations • 20 Nov 2017 • Ronak Mehta, Hyunwoo J. Kim, Shulei Wang, Sterling C. Johnson, Ming Yuan, Vikas Singh
Recent results in coupled or temporal graphical models offer schemes for estimating the relationship structure between features when the data come from related (but distinct) longitudinal sources.
no code implementations • ICCV 2017 • Rudrasis Chakraborty, Vikas Singh, Nagesh Adluru, Baba C. Vemuri
Finally, by using existing algorithms for recursive Frechet mean and exact principal geodesic analysis on the hypersphere, we present several experiments on synthetic and real (vision and medical) data sets showing how group testing on such diversely sampled longitudinal data is possible by analyzing the reconstructed data in the subspace spanned by the first few PGs.
1 code implementation • ICML 2017 • Hao Henry Zhou, Yilin Zhang, Vamsi K. Ithapu, Sterling C. Johnson, Grace Wahba, Vikas Singh
Many studies in biomedical and health sciences involve small sample sizes due to logistic or financial constraints.
no code implementations • 22 Aug 2017 • Sathya N. Ravi, Maxwell D. Collins, Vikas Singh
We present a new Frank-Wolfe (FW) type algorithm that is applicable to minimization problems with a nonsmooth convex objective.
1 code implementation • CVPR 2017 • Sathya N. Ravi, Yunyang Xiong, Lopamudra Mukherjee, Vikas Singh
This paper is inspired by a relatively recent work of Seitz and Baker which introduced the so-called Filter Flow model.
no code implementations • CVPR 2017 • Hyunwoo J. Kim, Nagesh Adluru, Heemanshu Suri, Baba C. Vemuri, Sterling C. Johnson, Vikas Singh
Statistical machine learning models that operate on manifold-valued data are being extensively studied in vision, motivated by applications in activity recognition, feature tracking and medical imaging.
no code implementations • CVPR 2017 • Vamsi K. Ithapu, Risi Kondor, Sterling C. Johnson, Vikas Singh
Multiresolution analysis and matrix factorization are foundational tools in computer vision.
no code implementations • 4 Mar 2017 • Felipe Gutierrez-Barragan, Vamsi K. Ithapu, Chris Hinrichs, Camille Maumet, Sterling C. Johnson, Thomas E. Nichols, Vikas Singh, the ADNI
We find that RapidPT achieves its best runtime performance on medium sized datasets ($50 \leq n \leq 200$), with speedups of 1. 5x - 38x (vs. SnPM13) and 20x-1000x (vs. NaivePT).
no code implementations • 28 Feb 2017 • Vamsi K. Ithapu, Sathya N. Ravi, Vikas Singh
We seek to analyze whether network architecture and input data statistics may guide the choices of learning parameters and vice versa.
no code implementations • NeurIPS 2016 • Hao Zhou, Vamsi K. Ithapu, Sathya Narayanan Ravi, Vikas Singh, Grace Wahba, Sterling C. Johnson
Consider samples from two different data sources $\{\mathbf{x_s^i}\} \sim P_{\rm source}$ and $\{\mathbf{x_t^i}\} \sim P_{\rm target}$.
no code implementations • CVPR 2016 • Won Hwa Kim, Hyunwoo J. Kim, Nagesh Adluru, Vikas Singh
A major goal of imaging studies such as the (ongoing) Human Connectome Project (HCP) is to characterize the structural network map of the human brain and identify its associations with covariates such as genotype, risk factors, and so on that correspond to an individual.
no code implementations • CVPR 2016 • Seong Jae Hwang, Nagesh Adluru, Maxwell D. Collins, Sathya N. Ravi, Barbara B. Bendlin, Sterling C. Johnson, Vikas Singh
There is a great deal of interest in using large scale brain imaging studies to understand how brain connectivity evolves over time for an individual and how it varies over different levels/quantiles of cognitive function.
no code implementations • ICCV 2015 • Won Hwa Kim, Sathya N. Ravi, Sterling C. Johnson, Ozioma C. Okonkwo, Vikas Singh
A variety of studies in neuroscience/neuroimaging seek to perform statistical inference on the acquired brain image scans for diagnosis as well as understanding the pathological manifestation of diseases.
no code implementations • ICCV 2015 • Seong Jae Hwang, Maxwell D. Collins, Sathya N. Ravi, Vamsi K. Ithapu, Nagesh Adluru, Sterling C. Johnson, Vikas Singh
Eigenvalue problems are ubiquitous in computer vision, covering a very broad spectrum of applications ranging from estimation problems in multi-view geometry to image segmentation.
no code implementations • ICCV 2015 • Hyunwoo J. Kim, Nagesh Adluru, Monami Banerjee, Baba C. Vemuri, Vikas Singh
Probability density functions (PDFs) are fundamental "objects" in mathematics with numerous applications in computer vision, machine learning and medical imaging.
no code implementations • ICCV 2015 • Lopamudra Mukherjee, Sathya N. Ravi, Vamsi K. Ithapu, Tyler Holmes, Vikas Singh
In this paper, we first derive an Augmented Lagrangian approach to optimize the standard binary Hashing objective (i. e., maintain fidelity with a given distance matrix).
no code implementations • 17 Nov 2015 • Vamsi K. Ithapu, Sathya N. Ravi, Vikas Singh
The regularization and output consistency behavior of dropout and layer-wise pretraining for learning deep networks have been fairly well studied.
no code implementations • 10 Jun 2015 • Vamsi K. Ithapu, Sathya Ravi, Vikas Singh
Unsupervised pretraining and dropout have been well studied, especially with respect to regularization and output consistency.
no code implementations • CVPR 2015 • Jia Xu, Lopamudra Mukherjee, Yin Li, Jamieson Warner, James M. Rehg, Vikas Singh
Motivated by these applications, this paper focuses on the problem of egocentric video summarization.
no code implementations • CVPR 2015 • Won Hwa Kim, Barbara B. Bendlin, Moo. K. Chung, Sterling C. Johnson, Vikas Singh
Statistical analysis of longitudinal or cross sectionalbrain imaging data to identify effects of neurodegenerative diseases is a fundamental task in various studies in neuroscience.
no code implementations • NeurIPS 2013 • Chris Hinrichs, Vamsi K. Ithapu, Qinyuan Sun, Sterling C. Johnson, Vikas Singh
In this paper, we show that permutation testing in fact amounts to populating the columns of a very large matrix ${\bf P}$.
no code implementations • 12 Feb 2015 • Vamsi K. Ithapu, Sathya Ravi, Vikas Singh
The success of deep architectures is at least in part attributed to the layer-by-layer unsupervised pre-training that initializes the network.
no code implementations • NeurIPS 2014 • Deepti Pachauri, Risi Kondor, Gautam Sargur, Vikas Singh
Consistently matching keypoints across images, and the related problem of finding clusters of nearby images, are critical components of various tasks in Computer Vision, including Structure from Motion (SfM).
no code implementations • CVPR 2014 • Hyunwoo J. Kim, Nagesh Adluru, Maxwell D. Collins, Moo. K. Chung, Barbara B. Bendlin, Sterling C. Johnson, Richard J. Davidson, Vikas Singh
Linear regression is a parametric model which is ubiquitous in scientific analysis.
no code implementations • NeurIPS 2013 • Deepti Pachauri, Risi Kondor, Vikas Singh
The problem of matching not just two, but m different sets of objects to each other arises in a variety of contexts, including finding the correspondence between feature points across multiple images in computer vision.
no code implementations • CVPR 2013 • Jia Xu, Maxwell D. Collins, Vikas Singh
We study the problem of interactive segmentation and contour completion for multiple objects.
no code implementations • CVPR 2013 • Won Hwa Kim, Moo. K. Chung, Vikas Singh
In this paper, we adapt recent results in harmonic analysis, to derive NonEuclidean Wavelets based algorithms for a range of shape analysis problems in vision and medical imaging.
no code implementations • NeurIPS 2012 • Won H. Kim, Deepti Pachauri, Charles Hatt, Moo. K. Chung, Sterling Johnson, Vikas Singh
In contrast to hypothesis tests on point-wise measurements, in this paper, we make the case for performing statistical analysis on multi-scale shape descriptors that characterize the local topological context of the signal around each surface vertex.
no code implementations • NeurIPS 2012 • Chris Hinrichs, Vikas Singh, Jiming Peng, Sterling Johnson
Multiple Kernel Learning (MKL) generalizes SVMs to the setting where one simultaneously trains a linear classifier and chooses an optimal combination of given base kernels.
no code implementations • NeurIPS 2010 • Kamiya Motwani, Nagesh Adluru, Chris Hinrichs, Andrew Alexander, Vikas Singh
Now, given such a representation, the problem reduces to segmenting new brain image with additional constraints that enforce consistency between the segmented foreground and the pre-specified histogram over features.