Welcome
I joined Meta as an Applied Research Scientist in the Surreal Team.
I completed my PhD at the University of Southern California (USC), Los Angeles under the supervision of Prof. Ram Nevatia. My primary field of research was at the intersection of Computer Vision and Natural Language Processing. Specifically, I focused on grounding language in vision, with an emphasis on videos.
As an undergraduate, I earned my degree from IIT Bombay. Throughout my academic journey, I have gained valuable experience through several internships including Meta AI (Summer’22), PRIOR@AI2 (Summer’20, 21), Wadhwani AI (Summer’18), USC (Summer’17), and Aalto University (Summer’16).
CV/Resume, LinkedIn, Google Scholar, Semantic Scholar, Github
Email ID: (ark.sadhu2904@gmail.com)
News and Updates
| June 2024 | I joined Meta in the Surreal Team. |
| May 2024 | I attended the hooding ceremony and officially submitted my thesis |
| April 2024 | I defended my thesis! The presentation slides can be found here. |
| Jan 2024 | I presented my paper Leveraging Task-Specific Pre-Training To Reason Across Images and Videos at WACV24. |
| 2023 | Served as a reviewer for ICML'23, ACL'23, ICCV'23, EMNLP'23, Neurips'23, BMVC'23, WACV'24, CVPR'24, AURO and TPAMI. |
| Nov 2022 | I will serve as a reviewer for CVPR'23. |
| Aug 2022 | I will serve as a reviewer for BMVC'22. |
| Aug 2022 | I will continue my internship with Meta AI as a part-time. |
| Jun 2022 | I will serve as a reviewer for WACV'23. |
| May 2022 | I started my internship at Meta AI in the AI commerce team with Animesh Sinha, Licheng Yu, Hugo Chen and Ning Zhang. |
| April 2022 | I was an highlighted reviewer for ICLR'22 [Link]. |
| March 2022 | I am serving as a reviewer for ECCV'22. |
| Nov 2021 | I am serving as a reviewer for ACL Rolling Review. |
| Nov 2021 | I will serve as a reviewer for CVPR'22. |
| Nov 2021 | I served as an emergency reviewer for ICLR'22. |
| Oct 2021 | I will be interning at Meta AI in the AI commerce team with Animesh Sinha, Licheng Yu, Hugo Chen and Ning Zhang. |
| Oct 2021 | I will serve as a reviewer for IJCV Journal. |
| Sep 2021 | I will serve as a reviewer for AAAI'22. |
| Jun 2021 | I will serve as a reviewer for BMVC'21. |
| Jun 2021 | Presented "Visual Semantic Role Labeling for Video Understanding" at CVPR'21. |
| Jun 2021 | Presented "Video Question Answering with Phrases via Semantic Roles" at NAACL'21. |
| May 2021 | I will serve as a reviewer for EMNLP'21. |
| May 2021 | Recognized as Outstanding Reviewer at CVPR'21 |
| May 2021 | Mentoring Pranit Chawla, an intern at our lab over Summer'21. |
| May 2021 | Started summer'21 internship in PRIOR team (AI2) with Ani Kembhavi and Tanmay Gupta. |
| April 2021 | Our work on "Video Question Answering with Phrases via Semantic Roles" (NAACL'21) is available on Arxiv, Github |
| April 2021 | VidSitu Challenge is a guest task at the ActivityNet Workshop at CVPR'21 this year! Checkout the challenge here. |
| April 2021 | Our work on "Visual Semantic Role Labeling for Video Understanding" (CVPR'21) is available on Arxiv with instructions on obtaining code, data and leaderboard on VidSitu Website. |
| March 2021 | I will serve as a reviewer for ICCV'21 |
| March 2021 | Our work on "Video Question Answering with Phrases via Semantic Roles" is accepted to NAACL'21. Paper and Code will be out soon! |
| March 2021 | Our work on "Visual Semantic Role Labeling for Video Understanding" has been accepted to CVPR'21. Arxiv,Code,Website coming soon. |
| Feb 2021 | I will serve as a Program Committee member at ALVR'21 workshop at NAACL'21. |
| Feb 2021 | I will serve as a reviewer for ACL'21 |
| Jan 2021 | I will be returning as an intern with the PRIOR Team (AI2) for an internship in Summer 2021. |
| Nov 2020 | I will serve as a reviewer for CVPR'21 |
| Nov 2020 | Our paper titled "Utilizing Every Image Object for Semi-supervised Phrase Grounding" has been accepted to WACV'21 and is available on Arxiv. |
| Sep 2020 | I will serve as a reviewer for AAAI'21 |
| Aug 2020 | Recognized as Outstanding Reviewer at BMVC'20 |
| June 2020 | Presented "Video Object Grounding using Semantic Roles in Language Description" at CVPR20. [5min Video]. |
| May 2020 | Started internship at AI2 in the Prior group with Ani Kembhavi, Tanmay Gupta, Mark Yatskar. |
| May 2020 | I will serve as a reviewer for BMVC20. |
| March 2020 | Our CVPR20 paper Video Object Grounding using Semantic Roles in Language Description is avaible on [Arxiv] and [Github] |
| March 2020 | Open-sourced a repository with research advices. |
| Feb 2020 | Our paper "Video Object Grounding using Semantic Roles in Language Description" has been accepted to CVRP20. Arxiv/Github coming soon. |
| Feb 2020 | I will serve as a reviewer for ICPR20. |
| Jan 2020 | I will join PRIOR Team (AI2) for an internship in Summer 2020. |
| Jan 2020 | Attended Google LA PhD summit. |
| Jan 2020 | Gave a talk on "Need for Language in Vision" as a part of USC-WiSE seminar. [Slides] |
| Nov 2019 | Attended AI symposium at USC. |
| Oct 2019 | Presented "Zero-Shot Grounding of Objects from Natural Language Queries" at ICCV 2019. [Video], [Slides], [Poster]. |
| Aug 2019 | Our paper "Zero-Shot Grounding of Objects from Natural Language Queries" is now available at Arxiv and Github. |
| Aug 2019 | I am serving as a reviewer for WACV'2020. |
| July 2019 | Our paper "Zero-Shot Grounding of Objects from Natural Language Queries" is accepted to ICCV'19 as an Oral paper. Arxiv, github code and a blog post coming soon. |
| June 2019 | Delivered a key-note at Media-Forensics workshop at CVPR'19 summarizing the key-takeaways from the Audio-Visual Fakes Workshop. [Slides] |
| June 2019 | Successfully co-organized the first iteration of the workshop Synthetic-Realities: Audio-Visual Fakes at ICML'19. Live Recording of the workshop is also available. |
| May 2019 | Mentoring Varun Sunar, a Viterbi-India fellow who will be interning at our lab over Summer'19. |
| April 2019 | Presented Rhythmic trajectories in Annenberg Symposium. Joint work with Szilvia Ruszev. |
| March 2019 | Co-organizing a workshop on detecting Audio-Visual Fakes at ICML'19 . The call for papers is out . Submission deadline is 1 May 2019 (anywhere on earth). |
| Jan 2019 | Released subreddit-classification dataset on Kaggle here. Also find corresponding github repository |
| Dec 2018 | Released VCR-Bert: Using BERT for QA on the VCR dataset. Achieves 6% more than the scores reported in the paper. |
| Oct 2018 | Released Awesome Grounding: A collection of papers on visual grounding which came out in the recent years. |
| Sep 2018 | I will be attending the Pytorch Dev Conference at San Fransisco on Oct 2nd. |
| Aug 2018 | I have joined University of Southern California as a PhD student. I am working in Prof. Ram Nevatia's lab in PHE234. |
| July 2018 | I have written a blog on AI courses offered at IITB. Check it out here |
| June 2018 | We placed third in the iFood Challenge. Our implementation is available here |
| April 2018 | I have accepted PhD offer at University of Southern California and I will start in Fall 2018 under Prof. Ram Nevatia |