Get our free extension to see links to code for papers anywhere online!

Add to Chrome

Add to Firefox

Picture for Alexey Romanov

Alexey Romanov

Lights, Camera, Action! A Framework to Improve NLP Accuracy over OCR documents

Aug 06, 2021
Amit Gupte, Alexey Romanov, Sahitya Mantravadi, Dalitso Banda, Jianjie Liu, Raza Khan, Lakshmanan Ramu Meenal, Benjamin Han, Soundar Srinivasan

Figure 1 for Lights, Camera, Action! A Framework to Improve NLP Accuracy over OCR documents

Figure 2 for Lights, Camera, Action! A Framework to Improve NLP Accuracy over OCR documents

Figure 3 for Lights, Camera, Action! A Framework to Improve NLP Accuracy over OCR documents

Figure 4 for Lights, Camera, Action! A Framework to Improve NLP Accuracy over OCR documents

* Accepted to the Document Intelligence Workshop at KDD 2021. The source code of Genalog is available at https://github.com/microsoft/genalog

Access Paper or Ask Questions

Revealing the Dark Secrets of BERT

Sep 11, 2019
Olga Kovaleva, Alexey Romanov, Anna Rogers, Anna Rumshisky

Figure 1 for Revealing the Dark Secrets of BERT

Figure 2 for Revealing the Dark Secrets of BERT

Figure 3 for Revealing the Dark Secrets of BERT

Figure 4 for Revealing the Dark Secrets of BERT

* Accepted to EMNLP 2019

Access Paper or Ask Questions

What's in a Name? Reducing Bias in Bios without Access to Protected Attributes

Apr 10, 2019
Alexey Romanov, Maria De-Arteaga, Hanna Wallach, Jennifer Chayes, Christian Borgs, Alexandra Chouldechova, Sahin Geyik, Krishnaram Kenthapadi, Anna Rumshisky, Adam Tauman Kalai

Figure 1 for What's in a Name? Reducing Bias in Bios without Access to Protected Attributes

Figure 2 for What's in a Name? Reducing Bias in Bios without Access to Protected Attributes

Figure 3 for What's in a Name? Reducing Bias in Bios without Access to Protected Attributes

Figure 4 for What's in a Name? Reducing Bias in Bios without Access to Protected Attributes

* Accepted at NAACL 2019; Best Thematic Paper

Access Paper or Ask Questions

Bias in Bios: A Case Study of Semantic Representation Bias in a High-Stakes Setting

Jan 27, 2019
Maria De-Arteaga, Alexey Romanov, Hanna Wallach, Jennifer Chayes, Christian Borgs, Alexandra Chouldechova, Sahin Geyik, Krishnaram Kenthapadi, Adam Tauman Kalai

Figure 1 for Bias in Bios: A Case Study of Semantic Representation Bias in a High-Stakes Setting

Figure 2 for Bias in Bios: A Case Study of Semantic Representation Bias in a High-Stakes Setting

Figure 3 for Bias in Bios: A Case Study of Semantic Representation Bias in a High-Stakes Setting

Figure 4 for Bias in Bios: A Case Study of Semantic Representation Bias in a High-Stakes Setting

* Accepted at ACM Conference on Fairness, Accountability, and Transparency (ACM FAT*), 2019

Access Paper or Ask Questions

Lessons from Natural Language Inference in the Clinical Domain

Aug 27, 2018
Alexey Romanov, Chaitanya Shivade

Figure 1 for Lessons from Natural Language Inference in the Clinical Domain

Figure 2 for Lessons from Natural Language Inference in the Clinical Domain

Figure 3 for Lessons from Natural Language Inference in the Clinical Domain

Figure 4 for Lessons from Natural Language Inference in the Clinical Domain

* Extended version of the EMNLP 2018 paper. Dataset and code available at https://jgc128.github.io/mednli/

Access Paper or Ask Questions

Here's My Point: Joint Pointer Architecture for Argument Mining

May 08, 2017
Peter Potash, Alexey Romanov, Anna Rumshisky

Figure 1 for Here's My Point: Joint Pointer Architecture for Argument Mining

Figure 2 for Here's My Point: Joint Pointer Architecture for Argument Mining

Figure 3 for Here's My Point: Joint Pointer Architecture for Argument Mining

Figure 4 for Here's My Point: Joint Pointer Architecture for Argument Mining

* 10 pages; under review for ICLR

Access Paper or Ask Questions

Forced to Learn: Discovering Disentangled Representations Without Exhaustive Labels

May 01, 2017
Alexey Romanov, Anna Rumshisky

Figure 1 for Forced to Learn: Discovering Disentangled Representations Without Exhaustive Labels

Figure 2 for Forced to Learn: Discovering Disentangled Representations Without Exhaustive Labels

Figure 3 for Forced to Learn: Discovering Disentangled Representations Without Exhaustive Labels

Figure 4 for Forced to Learn: Discovering Disentangled Representations Without Exhaustive Labels

* Abstract accepted at ICLR 2017 Workshop: https://openreview.net/pdf?id=SkCmfeSFg

Access Paper or Ask Questions