Research

An up-to-date list of my papers is available on Google Scholar. For publications loosely grouped by topics, see here.

Publications (by year)

2023

  1. arXiv
    Beyond First-Order Tweedie: Solving Inverse Problems using Latent Diffusion
    Litu Rout, Yujia Chen, Abhishek Kumar, Constantine Caramanis, Sanjay Shakkottai, and Wen-Sheng Chu
    arXiv preprint, 2023
    arXiv Website
  2. arXiv
    Small-scale proxies for large-scale Transformer training instabilities
    Mitchell Wortsman et. al
    arXiv preprint, 2023
    arXiv
  3. NeurIPS
    Towards last-layer retraining for group robustness with fewer annotations
    Tyler LaBonte, Vidya Muthukumar, and Abhishek Kumar
    In Advances in Neural Information Processing Systems (NeurIPS), 2023
    arXiv
  4. KDD
    To aggregate or not? learning with separate noisy labels
    Jiaheng Wei, Zhaowei Zhu, Tianyi Luo, Ehsan Amid, Abhishek Kumar, and Yang Liu
    In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 2023
    arXiv
  5. arXiv
    Understanding the Effectiveness of Early Weight Averaging for Training Large Language Models
    Sunny Sanyal, Jean Kaddour, Abhishek Kumar, and Sujay Sanghavi
    arXiv preprint, 2023
    arXiv
  6. arXiv
    Score-based causal representation learning with interventions
    Burak Varici, Emre Acarturk, Karthikeyan Shanmugam, Abhishek Kumar, and Ali Tajer
    arXiv preprint, 2023
    arXiv
  7. ICLR
    Distributionally robust posthoc classifiers under prior shift
    Jiaheng Wei, Harikrishna Narasimhan, Ehsan Amid, Vincent Chu, Yang Liu, and Abhishek Kumar
    In International Conference on Learning Representations (ICLR), 2023
    arXiv PDF

2022

  1. TMLR
    DiffuseVAE: Efficient, controllable and high-fidelity generation from low-dimensional latents
    Kushagra Pandey, Avideep Mukherjee, Piyush Rai, and Abhishek Kumar
    Transactions on Machine Learning Research, 2022
    arXiv Code
  2. NeurIPS-W
    Dropout Disagreement: A Recipe for Group Robustness with Fewer Annotations
    Tyler LaBonte, Vidya Muthukumar, and Abhishek Kumar
    In NeurIPS 2022 Workshop on Distribution Shifts, 2022
  3. NeurIPS-W
    Fast Implicit Constrained Optimization of Non-decomposable Objectives for Deep Networks
    Yatong Chen, Abhishek Kumar, Yang Liu, and Ehsan Amid
    In NeurIPS 2022 Workshop on ‘Has it Trained Yet?‘, 2022
    Website

2021

  1. ICML
    Implicit rate-constrained optimization of non-decomposable objectives
    Abhishek Kumar, Harikrishna Narasimhan, and Andrew Cotter
    In International Conference on Machine Learning (ICML), 2021
    arXiv Code
  2. ICLR Outstanding Paper
    Score-based generative modeling through stochastic differential equations
    Yang Song, Jascha Sohl-Dickstein, Diederik P Kingma, Abhishek Kumar, Stefano Ermon, and Ben Poole
    In International Conference on Learning Representations (ICLR), 2021
    arXiv Code
  3. arXiv
    Constrained Instance and Class Reweighting for Robust Learning under Label Noise
    Abhishek Kumar, and Ehsan Amid
    arXiv preprint, 2021
    arXiv Blog Code
  4. ICASSP
    A scale invariant measure of flatness for deep network minima
    Akshay Rangamani, Nam H Nguyen, Abhishek Kumar, Dzung Phan, Sang Peter Chin, and Trac D Tran
    In IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2021
    arXiv Code
  5. AAAI
    Generalized adversarially learned inference
    Yatin Dandi, Homanga Bharadhwaj, Abhishek Kumar, and Piyush Rai
    In AAAI, 2021
    arXiv
  6. MLHC
    Risk score learning for COVID-19 contact tracing apps
    Kevin Murphy, Abhishek Kumar, and Stylianos Serghiou
    In Machine Learning for Healthcare Conference (MLHC), 2021
    arXiv Code
  7. CVPR
    Retrieve in style: Unsupervised facial feature transfer and retrieval
    Min Jin Chong, Wen-Sheng Chu, Abhishek Kumar, and David Forsyth
    In IEEE/CVF International Conference on Computer Vision (CVPR), 2021
    arXiv Code
  8. arXiv
    Solving Inverse Problems with NerfGANs
    Giannis Daras, Wen-Sheng Chu, Abhishek Kumar, Dmitry Lagun, and Alexandros G Dimakis
    arXiv preprint, 2021
    arXiv

2020

  1. ICML
    On Implicit Regularization in β-VAEs
    Abhishek Kumar, and Ben Poole
    In International Conference on Machine Learning (ICML), 2020
    arXiv
  2. AISTATS
    Regularized autoencoders via relaxed injective probability flow
    Abhishek Kumar, Ben Poole, and Kevin Murphy
    In International Conference on Artificial Intelligence and Statistics (AISTATS), 2020
    arXiv
  3. ICLR
    Weakly supervised disentanglement with guarantees
    Rui Shu, Yining Chen, Abhishek Kumar, Stefano Ermon, and Ben Poole
    In International Conference on Learning Representations (ICLR), 2020
    arXiv
  4. arXiv
    Few-shot adaptation of generative adversarial networks
    Esther Robb, Wen-Sheng Chu, Abhishek Kumar, and Jia-Bin Huang
    arXiv preprint, 2020
    arXiv

2019

  1. CVPR
    Spottune: transfer learning through adaptive fine-tuning
    Yunhui Guo, Honghui Shi, Abhishek Kumar, Kristen Grauman, Tajana Rosing, and Rogerio Feris
    In IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019
    arXiv

2018

  1. ICLR
    Variational Inference of Disentangled Latent Concepts from Unlabeled Observations
    Abhishek Kumar, Prasanna Sattigeri, and Avinash Balakrishnan
    In International Conference on Learning Representations (ICLR), 2018
    arXiv PDF
  2. NeurIPS
    Co-regularized alignment for unsupervised domain adaptation
    Abhishek Kumar, Prasanna Sattigeri, Kahini Wadhawan, Leonid Karlinsky, Rogerio Feris, William T Freeman, and Gregory Wornell
    In Advances in Neural Information Processing Systems (NeurIPS), 2018
    arXiv
  3. CVPR-W Oral
    The riemannian geometry of deep generative models
    Hang Shao, Abhishek Kumar, and P Thomas Fletcher
    In 4th Intl. Workshop on Differential Geometry in Computer Vision and Machine Learning at CVPR, 2018
    arXiv Website
  4. CVPR Spotlight
    BlockDrop: Dynamic Inference Paths in Residual Networks
    Zuxuan Wu, Tushar Nagarajan, Abhishek Kumar, Steven Rennie, Larry S Davis, Kristen Grauman, and Rogerio Feris
    In IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018
    arXiv
  5. NeurIPS Spotlight
    Delta-encoder: an effective sample synthesis method for few-shot object recognition
    Eli Schwartz, Leonid Karlinsky, Joseph Shtok, Sivan Harary, Mattias Marder, Rogerio Feris, Abhishek Kumar, Raja Giryes, and Alex M Bronstein
    In Advances in Neural Information Processing Systems (NeurIPS), 2018
    arXiv
  6. arXiv
    Understanding unequal gender classification accuracy from face images
    Vidya Muthukumar, Tejaswini Pedapati, Nalini Ratha, Prasanna Sattigeri, Chai-Wah Wu, Brian Kingsbury, Abhishek Kumar, Samuel Thomas, Aleksandra Mojsilovic, and Kush R Varshney
    arXiv preprint, 2018
    arXiv

2017

  1. NeurIPS
    Semi-supervised Learning with GANs: Manifold Invariance with Improved Inference
    Abhishek Kumar, Prasanna Sattigeri, and Tom Fletcher
    In Advances in Neural Information Processing Systems (NeurIPS), 2017
    arXiv
  2. CVPR
    S3pool: Pooling with stochastic spatial sampling
    Shuangfei Zhai, Hui Wu, Abhishek Kumar, Yu Cheng, Yongxi Lu, Zhongfei Zhang, and Rogerio Feris
    In IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017
    arXiv Bib PDF
  3. AISTATS
    Local group invariant representations via orbit embeddings
    Anant Raj, Abhishek Kumar, Youssef Mroueh, P Thomas Fletcher, and B Scholkopf
    In International Conference on Artificial Intelligence and Statistics (AISTATS), 2017
    arXiv Website
  4. CVPR Spotlight
    Fully-adaptive feature sharing in multi-task networks with applications in person attribute classification
    Yongxi Lu, Abhishek Kumar, Shuangfei Zhai, Yu Cheng, Tara Javidi, and Rogerio Feris
    In IEEE conference on computer vision and pattern recognition (CVPR), 2017
    arXiv Bib PDF
  5. arXiv
    Sengen: Sentence generating neural variational topic model
    Ramesh Nallapati, Igor Melnyk, Abhishek Kumar, and Bowen Zhou
    arXiv preprint, 2017
    arXiv
  6. arXiv
    Improved neural text attribute transfer with non-parallel data
    Igor Melnyk, Cicero Nogueira dos Santos, Kahini Wadhawan, Inkit Padhi, and Abhishek Kumar
    arXiv preprint, 2017
    arXiv
  7. Journal
    Performance of natural language classifiers in a question-answering system
    Raimo Bakis, Daniel P Connors, Parijat Dube, Pavan Kapanipathi, Abhishek Kumar, Dmitry Malioutov, and Chitra Venkatramani
    IBM Journal of Research and Development, 2017

2016

  1. AISTATS
    Scalable exemplar clustering and facility location via augmented block coordinate descent with column generation
    Ian En-Hsu Yen, Dmitry Malioutov, and Abhishek Kumar
    In International Conference on Artificial Intelligence and Statistics (AISTATS), 2016
    Bib PDF Supp Website
  2. UAI
    Large-scale Submodular Greedy Exemplar Selection with Structured Similarity Matrices
    Dmitry Malioutov, Abhishek Kumar, and Ian En-Hsu Yen
    In Uncertainty in Artificial Intelligence (UAI), 2016
    PDF
  3. ACL Oral
    Incorporating relational knowledge into word representations using subspace regularization
    Abhishek Kumar, and Jun Araki
    In Proceedings of Association for Computational Linguistics, 2016
    Bib PDF Website

2015

  1. SDM Oral, Best of SDM
    Near-separable non-negative matrix factorization with ℓ1 and Bregman loss functions
    Abhishek Kumar, and Vikas Sindhwani
    In SIAM International Conference on Data Mining (SDM), 2015
    arXiv PDF
  2. SiMAX
    Exact and heuristic algorithms for semi-nonnegative matrix factorization
    Nicolas Gillis, and Abhishek Kumar
    SIAM Journal on Matrix Analysis and Applications, 2015
    arXiv Bib PDF

2013

  1. ICML Full Oral
    Fast Conical Hull Algorithms for Near-separable Non-negative Matrix Factorization
    Abhishek Kumar, Vikas Sindhwani, and Prabhanjan Kambadur
    In Proceedings of the 30th International Conference on Machine Learning (ICML), 2013
    arXiv Bib PDF

2012

  1. ICML Full Oral
    A Binary Classification Framework for Two-Stage Multiple Kernel Learning
    Abhishek Kumar, Alexandru Niculescu-Mizil, Koray Kavukcoglu, and Hal Daumé III
    In Proceedings of the 29th International Conference on Machine Learning (ICML), 2012
    Bib PDF Supp Code
  2. ICML Full Oral
    Learning task grouping and overlap in multi-task learning
    Abhishek Kumar, and Hal Daume III
    In International Coference on Machine Learning (ICML), 2012
    arXiv PDF
  3. CVPR
    Generalized multiview analysis: A discriminative latent space
    Abhishek Sharma, Abhishek Kumar, Hal Daume, and David W Jacobs
    In IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2012
    PDF
  4. NeurIPS
    Simultaneously leveraging output and task structures for multiple-output regression
    Piyush Rai, Abhishek Kumar, and Hal Daume
    In Advances in Neural Information Processing Systems (NeurIPS), 2012
    PDF Website
  5. MICCAI
    Mixed-effects shape models for estimating longitudinal changes in anatomy
    Manasi Datar, Prasanna Muralidharan, Abhishek Kumar, Sylvain Gouttard, Joseph Piven, Guido Gerig, Ross Whitaker, and P Thomas Fletcher
    In International MICCAI Workshop on Spatio-temporal Image Analysis for Longitudinal and Time-Series Image Data, 2012
    Bib PDF

2011

  1. ICML
    A co-training approach for multi-view spectral clustering
    Abhishek Kumar, and Hal Daumé
    In International Conference on Machine Learning (ICML), 2011
    PDF Code
  2. NeurIPS
    Co-regularized multi-view spectral clustering
    Abhishek Kumar, Piyush Rai, and Hal Daume
    In Advances in neural information processing systems (NeurIPS), 2011
    PDF Code Website

2010

  1. M.S. Thesis
    Environment mismatch compensation methods for robust speech recognition
    Abhishek Kumar
    2010
    Bib
  2. ACL-W
    Frustratingly easy semi-supervised domain adaptation
    Hal Daumé III, Abhishek Kumar, and Avishek Saha
    In Proceedings of the 2010 ACL Workshop on Domain Adaptation for Natural Language Processing, 2010
    Bib PDF Website
  3. NeurIPS
    Co-regularization based semi-supervised domain adaptation
    Abhishek Kumar, Avishek Saha, and Hal Daume
    In Advances in Neural Information Processing Systems (NeurIPS), 2010
    Bib PDF Supp Website
  4. NeurIPS-W
    Co-regularized spectral clustering with multiple kernels
    Abhishek Kumar, Piyush Rai, and Hal Daumé III
    In NeurIPS 2010 workshop: New directions in multiple kernel learning, 2010
    PDF Website
  5. ICASSP
    A kernel mean matching approach for environment mismatch compensation in speech recognition
    Abhishek Kumar, and John HL Hansen
    In IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2010
    Bib

2009

  1. Report
    Project Report: Analysis of JL Transform Implementations
    Abhishek Kumar, and Manasi Datar
    , 2009
    Bib

2008

  1. Interspeech
    Environment mismatch compensation using average eigenspace for speech recognition
    Abhishek Kumar, and John HL Hansen
    In Interspeech, 2008
    Bib PDF

Publications (by topic)

Distribution or task shift

  1. NeurIPS
    Towards last-layer retraining for group robustness with fewer annotations
    Tyler LaBonte, Vidya Muthukumar, and Abhishek Kumar
    In Advances in Neural Information Processing Systems (NeurIPS), 2023
    arXiv
  2. KDD
    To aggregate or not? learning with separate noisy labels
    Jiaheng Wei, Zhaowei Zhu, Tianyi Luo, Ehsan Amid, Abhishek Kumar, and Yang Liu
    In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 2023
    arXiv
  3. ICLR
    Distributionally robust posthoc classifiers under prior shift
    Jiaheng Wei, Harikrishna Narasimhan, Ehsan Amid, Vincent Chu, Yang Liu, and Abhishek Kumar
    In International Conference on Learning Representations (ICLR), 2023
    arXiv PDF
  4. NeurIPS-W
    Dropout Disagreement: A Recipe for Group Robustness with Fewer Annotations
    Tyler LaBonte, Vidya Muthukumar, and Abhishek Kumar
    In NeurIPS 2022 Workshop on Distribution Shifts, 2022
  5. arXiv
    Constrained Instance and Class Reweighting for Robust Learning under Label Noise
    Abhishek Kumar, and Ehsan Amid
    arXiv preprint, 2021
    arXiv Blog Code
  6. CVPR
    Spottune: transfer learning through adaptive fine-tuning
    Yunhui Guo, Honghui Shi, Abhishek Kumar, Kristen Grauman, Tajana Rosing, and Rogerio Feris
    In IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019
    arXiv
  7. NeurIPS
    Co-regularized alignment for unsupervised domain adaptation
    Abhishek Kumar, Prasanna Sattigeri, Kahini Wadhawan, Leonid Karlinsky, Rogerio Feris, William T Freeman, and Gregory Wornell
    In Advances in Neural Information Processing Systems (NeurIPS), 2018
    arXiv
  8. arXiv
    Understanding unequal gender classification accuracy from face images
    Vidya Muthukumar, Tejaswini Pedapati, Nalini Ratha, Prasanna Sattigeri, Chai-Wah Wu, Brian Kingsbury, Abhishek Kumar, Samuel Thomas, Aleksandra Mojsilovic, and Kush R Varshney
    arXiv preprint, 2018
    arXiv
  9. M.S. Thesis
    Environment mismatch compensation methods for robust speech recognition
    Abhishek Kumar
    2010
    Bib
  10. ACL-W
    Frustratingly easy semi-supervised domain adaptation
    Hal Daumé III, Abhishek Kumar, and Avishek Saha
    In Proceedings of the 2010 ACL Workshop on Domain Adaptation for Natural Language Processing, 2010
    Bib PDF Website
  11. NeurIPS
    Co-regularization based semi-supervised domain adaptation
    Abhishek Kumar, Avishek Saha, and Hal Daume
    In Advances in Neural Information Processing Systems (NeurIPS), 2010
    Bib PDF Supp Website
  12. ICASSP
    A kernel mean matching approach for environment mismatch compensation in speech recognition
    Abhishek Kumar, and John HL Hansen
    In IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2010
    Bib
  13. Interspeech
    Environment mismatch compensation using average eigenspace for speech recognition
    Abhishek Kumar, and John HL Hansen
    In Interspeech, 2008
    Bib PDF

Generative models

  1. arXiv
    Beyond First-Order Tweedie: Solving Inverse Problems using Latent Diffusion
    Litu Rout, Yujia Chen, Abhishek Kumar, Constantine Caramanis, Sanjay Shakkottai, and Wen-Sheng Chu
    arXiv preprint, 2023
    arXiv Website
  2. arXiv
    Score-based causal representation learning with interventions
    Burak Varici, Emre Acarturk, Karthikeyan Shanmugam, Abhishek Kumar, and Ali Tajer
    arXiv preprint, 2023
    arXiv
  3. TMLR
    DiffuseVAE: Efficient, controllable and high-fidelity generation from low-dimensional latents
    Kushagra Pandey, Avideep Mukherjee, Piyush Rai, and Abhishek Kumar
    Transactions on Machine Learning Research, 2022
    arXiv Code
  4. ICLR Outstanding Paper
    Score-based generative modeling through stochastic differential equations
    Yang Song, Jascha Sohl-Dickstein, Diederik P Kingma, Abhishek Kumar, Stefano Ermon, and Ben Poole
    In International Conference on Learning Representations (ICLR), 2021
    arXiv Code
  5. AAAI
    Generalized adversarially learned inference
    Yatin Dandi, Homanga Bharadhwaj, Abhishek Kumar, and Piyush Rai
    In AAAI, 2021
    arXiv
  6. CVPR
    Retrieve in style: Unsupervised facial feature transfer and retrieval
    Min Jin Chong, Wen-Sheng Chu, Abhishek Kumar, and David Forsyth
    In IEEE/CVF International Conference on Computer Vision (CVPR), 2021
    arXiv Code
  7. arXiv
    Solving Inverse Problems with NerfGANs
    Giannis Daras, Wen-Sheng Chu, Abhishek Kumar, Dmitry Lagun, and Alexandros G Dimakis
    arXiv preprint, 2021
    arXiv
  8. ICML
    On Implicit Regularization in β-VAEs
    Abhishek Kumar, and Ben Poole
    In International Conference on Machine Learning (ICML), 2020
    arXiv
  9. AISTATS
    Regularized autoencoders via relaxed injective probability flow
    Abhishek Kumar, Ben Poole, and Kevin Murphy
    In International Conference on Artificial Intelligence and Statistics (AISTATS), 2020
    arXiv
  10. ICLR
    Weakly supervised disentanglement with guarantees
    Rui Shu, Yining Chen, Abhishek Kumar, Stefano Ermon, and Ben Poole
    In International Conference on Learning Representations (ICLR), 2020
    arXiv
  11. arXiv
    Few-shot adaptation of generative adversarial networks
    Esther Robb, Wen-Sheng Chu, Abhishek Kumar, and Jia-Bin Huang
    arXiv preprint, 2020
    arXiv
  12. ICLR
    Variational Inference of Disentangled Latent Concepts from Unlabeled Observations
    Abhishek Kumar, Prasanna Sattigeri, and Avinash Balakrishnan
    In International Conference on Learning Representations (ICLR), 2018
    arXiv PDF
  13. CVPR-W Oral
    The riemannian geometry of deep generative models
    Hang Shao, Abhishek Kumar, and P Thomas Fletcher
    In 4th Intl. Workshop on Differential Geometry in Computer Vision and Machine Learning at CVPR, 2018
    arXiv Website

Matrix factorization

  1. SDM Oral, Best of SDM
    Near-separable non-negative matrix factorization with ℓ1 and Bregman loss functions
    Abhishek Kumar, and Vikas Sindhwani
    In SIAM International Conference on Data Mining (SDM), 2015
    arXiv PDF
  2. SiMAX
    Exact and heuristic algorithms for semi-nonnegative matrix factorization
    Nicolas Gillis, and Abhishek Kumar
    SIAM Journal on Matrix Analysis and Applications, 2015
    arXiv Bib PDF
  3. ICML Full Oral
    Fast Conical Hull Algorithms for Near-separable Non-negative Matrix Factorization
    Abhishek Kumar, Vikas Sindhwani, and Prabhanjan Kambadur
    In Proceedings of the 30th International Conference on Machine Learning (ICML), 2013
    arXiv Bib PDF

Inducive bias

  1. NeurIPS
    Semi-supervised Learning with GANs: Manifold Invariance with Improved Inference
    Abhishek Kumar, Prasanna Sattigeri, and Tom Fletcher
    In Advances in Neural Information Processing Systems (NeurIPS), 2017
    arXiv
  2. AISTATS
    Local group invariant representations via orbit embeddings
    Anant Raj, Abhishek Kumar, Youssef Mroueh, P Thomas Fletcher, and B Scholkopf
    In International Conference on Artificial Intelligence and Statistics (AISTATS), 2017
    arXiv Website

Multitask learning

  1. CVPR Spotlight
    Fully-adaptive feature sharing in multi-task networks with applications in person attribute classification
    Yongxi Lu, Abhishek Kumar, Shuangfei Zhai, Yu Cheng, Tara Javidi, and Rogerio Feris
    In IEEE conference on computer vision and pattern recognition (CVPR), 2017
    arXiv Bib PDF
  2. ICML Full Oral
    Learning task grouping and overlap in multi-task learning
    Abhishek Kumar, and Hal Daume III
    In International Coference on Machine Learning (ICML), 2012
    arXiv PDF
  3. NeurIPS
    Simultaneously leveraging output and task structures for multiple-output regression
    Piyush Rai, Abhishek Kumar, and Hal Daume
    In Advances in Neural Information Processing Systems (NeurIPS), 2012
    PDF Website

Multi-modal learning

  1. ICML Full Oral
    A Binary Classification Framework for Two-Stage Multiple Kernel Learning
    Abhishek Kumar, Alexandru Niculescu-Mizil, Koray Kavukcoglu, and Hal Daumé III
    In Proceedings of the 29th International Conference on Machine Learning (ICML), 2012
    Bib PDF Supp Code
  2. CVPR
    Generalized multiview analysis: A discriminative latent space
    Abhishek Sharma, Abhishek Kumar, Hal Daume, and David W Jacobs
    In IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2012
    PDF
  3. ICML
    A co-training approach for multi-view spectral clustering
    Abhishek Kumar, and Hal Daumé
    In International Conference on Machine Learning (ICML), 2011
    PDF Code
  4. NeurIPS
    Co-regularized multi-view spectral clustering
    Abhishek Kumar, Piyush Rai, and Hal Daume
    In Advances in neural information processing systems (NeurIPS), 2011
    PDF Code Website
  5. NeurIPS-W
    Co-regularized spectral clustering with multiple kernels
    Abhishek Kumar, Piyush Rai, and Hal Daumé III
    In NeurIPS 2010 workshop: New directions in multiple kernel learning, 2010
    PDF Website

Deep learning (other)

  1. arXiv
    Small-scale proxies for large-scale Transformer training instabilities
    Mitchell Wortsman et. al
    arXiv preprint, 2023
    arXiv
  2. NeurIPS-W
    Fast Implicit Constrained Optimization of Non-decomposable Objectives for Deep Networks
    Yatong Chen, Abhishek Kumar, Yang Liu, and Ehsan Amid
    In NeurIPS 2022 Workshop on ‘Has it Trained Yet?‘, 2022
    Website
  3. ICML
    Implicit rate-constrained optimization of non-decomposable objectives
    Abhishek Kumar, Harikrishna Narasimhan, and Andrew Cotter
    In International Conference on Machine Learning (ICML), 2021
    arXiv Code
  4. ICASSP
    A scale invariant measure of flatness for deep network minima
    Akshay Rangamani, Nam H Nguyen, Abhishek Kumar, Dzung Phan, Sang Peter Chin, and Trac D Tran
    In IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2021
    arXiv Code
  5. CVPR Spotlight
    BlockDrop: Dynamic Inference Paths in Residual Networks
    Zuxuan Wu, Tushar Nagarajan, Abhishek Kumar, Steven Rennie, Larry S Davis, Kristen Grauman, and Rogerio Feris
    In IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018
    arXiv
  6. NeurIPS Spotlight
    Delta-encoder: an effective sample synthesis method for few-shot object recognition
    Eli Schwartz, Leonid Karlinsky, Joseph Shtok, Sivan Harary, Mattias Marder, Rogerio Feris, Abhishek Kumar, Raja Giryes, and Alex M Bronstein
    In Advances in Neural Information Processing Systems (NeurIPS), 2018
    arXiv
  7. CVPR
    S3pool: Pooling with stochastic spatial sampling
    Shuangfei Zhai, Hui Wu, Abhishek Kumar, Yu Cheng, Yongxi Lu, Zhongfei Zhang, and Rogerio Feris
    In IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017
    arXiv Bib PDF
  8. arXiv
    Sengen: Sentence generating neural variational topic model
    Ramesh Nallapati, Igor Melnyk, Abhishek Kumar, and Bowen Zhou
    arXiv preprint, 2017
    arXiv
  9. arXiv
    Improved neural text attribute transfer with non-parallel data
    Igor Melnyk, Cicero Nogueira dos Santos, Kahini Wadhawan, Inkit Padhi, and Abhishek Kumar
    arXiv preprint, 2017
    arXiv