Arjun Akula (Arjun Reddy Akula)

PhD Candidate
Center for Vision, Cognition, Learning, and Autonomy
University of California, Los Angeles (UCLA)

Email: aakula-at-ucla.edu
Previous: Google Research; Amazon Alexa AI; IBM Research; IIIT Hyderabad [Curriculum Vitae] [Google Scholar]


About

I am currently a 5th year PhD candidate at UCLA jointly supervised by Prof. Song-Chun Zhu (UCLA) and Prof. Joyce Chai (UMich). My research interests are in computer vision, natural language processing (NLP), statistical modeling and inference, and deep learning, with the focuses on interpretability, robustness, and trust in vision & language grounding models.

I also work closely with Prof. Siva Reddy (Mila & Stanford), Prof. Sinisa Todorovic (Oregon State University), Dr. Spandana Gella (Amazon AI & University of Edinburgh), Dr. Xiaodan Liang (CMU), Dr. Varun Jampani (Google Research), and Dr. Changsong Liu (UCLA & DMAI).

Prior to this, I was a Research Software Engineer at IBM Research in the Cognitive Research Department supervised by Dr. Gargi B Dasgupta. I did my Bachelors and Masters (by Research) in Computer Science and Engineering at IIIT Hyderabad, India. For my masters thesis, I worked on Question Answering Systems and Dialogue Systems at Language Technologies Research Center (LTRC) under the supervision of Prof. Radhika Mamidi and Prof. Rajeev Sangal.

Academic Activities
Sub-Reviewer: CVPR 2019EMNLP-IJCNLP 2019ACL 2019ECCV 2018EMNLP 2018EMNLP 2017
Panelist/Organizer: UCLA Data Science Workshop 2018, UCLA ASA DataFest 2017
Graduate Admission Review: CS Department, UCLA, 2020

News

Selected Publications

CrossVQA: Scalably Generating Benchmarks for Systematically Testing VQA Generalization
Arjun Akula, Soravit Changpinyo, Boqing Gong, Piyush Sharma, Song-Chun Zhu and Radu Soricut.
EMNLP 2021 (Long Paper, Main)
More details coming soon ...
Mind the Context: The Impact of Contextualization in Neural Module Networks for Grounding Visual Referring Expressions
Arjun Akula, Spandana Gella, Keze Wang, Song-Chun Zhu and Siva Reddy
EMNLP 2021 (Long Paper, Main)
More details coming soon ...
X-ToM: Explaining with Theory-of-Mind for Gaining Justified Human Trust
Arjun R Akula, Keze Wang, Changsong Liu, Sari Saba-Sadiya, Hongjing Lu, Sinisa Todorovic, Joyce Y Chai, Song-Chun Zhu.
iScience Cell Press 2021 (arXiv:1909.06907).
(A short version was presented at CVPRW 2019).
[abs] [pdf] [code]
CoCoX: Generating Conceptual and Counterfactual Explanations via Fault-Lines
Arjun R Akula, Shuai Wang, Song-Chun Zhu.
AAAI 2020
(acceptance rate: 20.5%) [Oral, Spotlight]
[abs] [pdf] [code][poster][slides]
Words aren't enough, their order matters: On the Robustness of Grounding Visual Referring Expressions
Arjun R Akula, Spandana Gella, Yaser Al-Onaizan, Song-Chun Zhu, Siva Reddy.
ACL 2020
(acceptane rate: 17.6%) [Oral Presentation]
[abs] [data] [pdf] [code] [slides] [poster]
Visual Discourse Parsing
Arjun R Akula, Song-Chun Zhu.
CVPR 2019 workshop on Language and Vision (arXiv:1903.02252v1) [Oral].
Also, selected as one among the three best oral papers.
[abs] [pdf] [code] [slides] [poster]
Explainable AI as Collaborative Task Solving
Arjun R Akula, Changsong Liu, Sinisa Todorovic, Joyce Y Chai, Song-Chun Zhu.
CVPR 2019 workshop on Explainable AI
[abs] [pdf] [code] [poster]
Natural Language Interaction with Explainable AI Models
Arjun R Akula, Sinisa Todorovic, Joyce Y Chai, Song-Chun Zhu.
CVPR 2019 workshop on Explainable AI
[abs] [pdf] [code]
Automatic problem extraction and analysis from unstructured text in IT tickets
S Agarwal, V Aggarwal, Arjun Akula, Gargi Dasgupta, G Sridhara
IBM Journal of Research and Development 2017
[abs] [pdf]
Classification of Attributes in a Natural Language Query into Different SQL Clauses
Ashish Palakurthi, Ruthu S.M., Arjun Akula, and Radhika Mamidi.
Recent Advances in Natural Language Processing (RANLP 2015)
[abs] [pdf]
Towards Auto-Remediation in Services Delivery: Context-based Classification of Noisy and Unstructured Tickets
Gargi Dasgupta, Tapan Nayak, Arjun Akula, Shivali Agarwal, Shripad Nadgowda.
International Conference on Service-Oriented Computing (ICSOC 2014)
[abs] [pdf]
A Novel Approach towards Incorporating Context Processing Capabilities in NLIDB system
Arjun Akula, Rajeev Sangal and Radhika Mamidi.
International Joint Conference on Natural Language Processing (IJCNLP 2013)
[abs] [pdf]
A Novel Approach towards Building a Portable NLIDB system using the CPG Framework
Abhijeet Gupta, Arjun Akula, Deepak Malladi, Puneeth Kukkadapu, Vinay Ainavolu and Rajeev Sangal.
International Conference on Asian Language Processing (IALP 2012)
[abs] [pdf]

Professional Experience


Applied Scientist Intern at Amazon Alexa AI (Summer 2021)
Worked on Vision-Language-Navigation models for ALFRED benchmark.
Mentors: Dr. Spandana Gella, Prof. Jesse Thomason, Prof. Mohit Bansal, Dr. Dilek Hakkani-Tur.

Research Intern at Google Research (Summer 2020)
Worked on quantifying cross-dataset distribution shifts in VQA.
Mentors: Dr. Soravit Changpinyo, Dr. Radu Soricut.

Applied Scientist Intern at Amazon AI (Summer 2019)
Worked on improving robustness of visual referring expression grounding models.
Mentors: Dr. Spandana Gella, Prof. Siva Reddy, Dr. Yaser Al Onaizan.

Research Collaboration with Mila (Spring & Winter 2020)
Worked on contextualizing neural module networks.
Mentors: Dr. Spandana Gella, Prof. Siva Reddy

Research Software Engineer at IBM Research (2014 - 2016)
Worked on human language technologies, IBM Watson QA system.

Patents

Analyzing Unstructured Ticket Text Using Discourse Cues in Communication Logs
Arjun Akula, Gargi B Dasgupta, Tapan K Nayak
Disclosure Number: IN920150227, July 2015. US Patent App.
A System and Method for Structured Representation and Classification of Unstructured Tickets in Services Delivery
Shivali A, Arjun Akula, Gargi B, Tapan K, Shripad J.
Disclosure Number: IN820140677, Oct 2014. US Patent App.
Measuring Effective Utilization of a Service Practitioner for Ticket Resolution via a Wearable Device
Arjun Akula, Gargi Dasgupta, Vijay Ekambaram, Ramasuri Narayanam.
Disclosure Number: IN920160178US1, Aug 2016. US Patent App

Education

University of California, Los Angeles (UCLA)
Sep. 2016 -- Present
PhD candidate
Advisor: Prof. Song-Chun Zhu, Prof. Joyce Chai
IIIT Hyderabad, India
Jun. 2012 -- Mar. 2014
Masters (by Research/Thesis) in Computer Science and Engineering
Advisor: Prof. Radhika Mamidi, Prof. Rajeev Sangal
IIIT Hyderabad, India
Aug. 2008 -- Jun. 2012
Bachelors in Computer Science and Engineering

Awards

  • Research Award - 2012 at IIIT Hyderabad for outstanding research work on Question Answering Systems.
  • Research Division Award - 2016 at IBM Research for proposing a novel text classification algorithm to classify problem tickets in IT Incident Management Process.
  • Outstanding Research Accomplishment Award - 2015 at IBM Research for designing and developing a web based reporting tool to measure adoption, utilization and business benefits of dynamic automations in IT Incident Management Process.

Invited Talks

Research Talk on Interpretability, Robustness, and Trust in Vision and Language Grounding Models
Google Research, Feb 2021.
Research Talk on Interpretability, Robustness, and Trust in Vision and Language Grounding Models
Microsoft AI, Feb 2021.
Research Talk on Interpretability, Robustness, and Trust in Vision and Language Grounding Models
Baidu Research, Oct 2020.
Research Talk on Interpretability, Robustness, and Trust in Vision and Language Grounding Models
Adobe Research, Oct 2020.
Research Talk on Robustness of Grounding Visual Referring Expression
Google AI, Aug 2020.
Research Talk on Interpretability and Trust in Vision and Language Grounding Models
UCLA, Aug 2019.
Research Talk on Semantic Role Labelling for Watson Question Answering System
IBM Research, Feb 2016.
Research Talk on Context based Natural Language Interfaces to Databases (NLIDB) and Dialogue Systems
Sixth IIIT-H Advanced Summer School on NLP (IASNLP 2015).