UMD Computer Science Researchers to Present 23 Papers at Major International ML and AI Events

UMDCS researchers will present papers at premier, renowned global conferences focused on Machine Learning and Artificial intelligence.

The International Conference on Machine Learning (ICML) (July 12 to July 18, 2020) is a globally renowned venue for presenting and publishing cutting-edge research on all aspects of machine learning used in closely related areas like artificial intelligence, statistics and data science, as well as important application areas such as machine vision, computational biology, speech recognition, and robotics. 

 UMD researchers will present 14 papers in the main track of this conference, as well as many working papers at workshops.  UMD is also heavily involved in the organization of this conference and its constituent workshops; indeed, UMD Professor Hal Daumé III is serving as Program Chair of ICML, jointly with CMU Professor Aarti Singh.

The International Conference on Learning Representations (ICLR) is dedicated to the advancement of the branch of artificial intelligence called representation learning, but generally referred to as deep learning.  ICLR is globally renowned for presenting and publishing cutting-edge research on all aspects of deep learning used in the fields of artificial intelligence, statistics and data science, as well as important application areas such as machine vision, computational biology, speech recognition, text understanding, gaming, and robotics.

The International Joint Conference on Artificial Intelligence (IJCAI) (originally scheduled July 11th to July 17th, 2020) is one of the oldest AI conferences in the world, having run continuously since 1969.  It’s publication footprint broadly spans AI, ML, and applications of techniques from those areas to societal implications.

We list papers to be presented at ICML, ICLR, and IJCAI below.

ICML

Scalable Differentiable Physics for Learning and Control

Yi-Ling Qiao (University of Maryland, College Park) · Junbang Liang (University of Maryland, College Park) · Vladlen Koltun (Intel Labs) · Ming Lin (UMD-CP & UNC-CH )

DRWR: A Differentiable Renderer without Rendering for Unsupervised 3D Structure Learning from Silhouette Images

Zhizhong Han (University of Maryland, College Park) · Chao Chen (Tsinghua University) · Yu-Shen Liu (Tsinghua University) · Matthias Zwicker (University of Maryland)

Certified Data Removal from Machine Learning Models

Chuan Guo (Cornell University) · Tom Goldstein (University of Maryland) · Awni Hannun (Facebook AI Research) · Laurens van der Maaten (Facebook)

Measuring Non-Expert Comprehension of Machine Learning Fairness Metrics

Debjani Saha (University of Maryland) · Candice Schumann (University of Maryland) · Duncan McElfresh (University of Maryland) · John P Dickerson (University of Maryland) · Michelle Mazurek (University of Maryland) · Michael Tschantz (International Computer Science Institute)

Adversarial Attacks on Copyright Detection Systems

Parsa Saadatpanah (University of Maryland) · Ali Shafahi (University of Maryland) · Tom Goldstein (University of Maryland)

Frequency Bias in Neural Networks for Input of Non-Uniform Density

Ronen Basri (Weizmann Institute of Science) · Meirav Galun (Weizmann Institute of Science) · Amnon Geifman (Weizmann Institute) · David Jacobs (University of Maryland, USA) · Yoni Kasten (Weizmann Institute) · Shira Kritchman (Weizmann Institute)

An end-to-end Differentially Private Latent Dirichlet Allocation Using a Spectral Algorithm

Christopher R DeCarolis (University of Maryland) · Mukul A Ram (University of Maryland) · Seyed Esmaeili (University of Maryland, College Park) · Yu-Xiang Wang (UC Santa Barbara) · Furong Huang (University of Maryland College Park)

Second-Order Provable Defenses against Adversarial Attacks

Sahil Singla (University of Maryland) · Soheil Feizi (University of Maryland)

Curse of Dimensionality on Randomized Smoothing for Certifiable Robustness

Aounon Kumar (University of Maryland, College Park) · Alexander Levine (University of Maryland) · Tom Goldstein (University of Maryland) · Soheil Feizi (University of Maryland)

On Second-Order Group Influence Functions for Black-Box Predictions

Samyadeep Basu (UMD) · Xuchen You (University of Maryland) · Soheil Feizi (University of Maryland)

Unraveling Meta-Learning: Understanding Feature Representations for Few-Shot Tasks

Micah Goldblum (University of Maryland) · Liam Fowl (University of Maryland) · Renkun Ni (University of Maryland) · Steven Reich (University of Maryland) · Valeriia Cherepanova (University of Maryland) · Tom Goldstein (University of Maryland)

A Pairwise Fair and Community-preserving Approach to k-Center Clustering

Brian Brubach (University of Maryland) · Darshan Chakrabarti (Carnegie Mellon University) · John P Dickerson (University of Maryland) · Samir Khuller (Northwestern University) · Aravind Srinivasan (University of Maryland College Park) · Leonidas Tsepenekas (University of Maryland, College Park)

Analyzing the effect of neural network architecture on training performance

Karthik Abinav Sankararaman (Facebook) · Soham De (DeepMind) · Zheng Xu (University of Maryland) · W. Ronny Huang (University of Maryland and EY LLP) · Tom Goldstein (University of Maryland)

ICLR

Breaking Certified Defenses: Semantic Adversarial Examples with Spoofed Robustness Certificates

Amin Ghiasi, Ali Shafahi, Tom Goldstein

FreeLB: Enhanced Adversarial Training for Natural Language Understanding

Chen Zhu, Yu Cheng, Zhe Gan, Siqi Sun, Tom Goldstein, Jingjing Liu

Network Deconvolution

Chengxi Ye, Matthew Evanusa, Hua He, Anton Mitrokhin, Tom Goldstein, James A. Yorke, Cornelia Fermuller, Yiannis Aloimonos- 

Certified Defenses for Adversarial Patches

Ping-yeh Chiang, Renkun Ni, Ahmed Abdelkader, Chen Zhu, Christoph Studor, Tom Goldstein

Adversarially robust transfer learning

Ali Shafahi, Parsa Saadatpanah, Chen Zhu, Amin Ghiasi, Christoph Studer, David Jacobs, Tom Goldstein

Truth or backpropaganda? An empirical investigation of deep learning theory

Micah Goldblum, Jonas Geiping, Avi Schwarzschild, Michael Moeller, Tom Goldstein

Sampling-Free Learning of Bayesian Quantized Neural Networks

Jiahao Su, Milan Cvitkovic, Furong Huang

Scalable Model Compression by Entropy Penalized Reparameterization

Deniz Oktay, Johannes Ballé, Saurabh Singh, Abhinav Shrivastava- 

IJCAI

An Algorithm for Multi-Attribute Diverse Matching

Saba Ahmadi (UMD), Faez Ahmed (Northwestern), John P. Dickerson (UMD), Mark Fuge (UMD), Samir Khuller (Northwestern)

Crowd-Steer: Realtime Smooth and Collision-Free Robot Navigation in Densely Crowded Scenarios Trained using High-Fidelity Simulation

Jing Liang (UMD), Utsav Patel (UMD), Adarsh Jagan Sathyamoorthy (UMD), Dinesh Manocha (UMD)