I am a research scientist at FAIR, Meta. I graduated with a PhD in Computer Science from UT Austin, advised by Dr. Kristen Grauman. My PhD research was at the intersection of video understanding and embodied AI.

Before coming to UT, I was an intern at MALL Lab, IISc, working with Dr. Partha Talukdar. I completed my B.E. in Computer Science from BITS Goa.

Contact: tushar.nagarajan@utexas.edu | tusharn@meta.com
CV: Link

News & Activities

Feb 24: We are organizing the First Joint Egocentric Vision (EgoVis) Workshop at CVPR 24 (17th June). [website]
Feb 24: Four papers accepted at CVPR 24: StepDiff, Video Detours, Video ReCap and Ego-Exo4D.
Nov 23: Ego-Exo4D has been announced! [arxiv] [blog] [website].
Sep 23: Two papers accepted at NeurIPS 23 (Datasets and benchmarks): Ego4D-GoalStep and HT-Step.
Sep 23: Two papers accepted at NeurIPS 23: EgoDistill and EgoEnv.
Nov 22: I graduated! Next stop: FAIR, Meta in NY. [thesis] [slides]
Mar 22: Ego4D has been accepted at CVPR 2022. [paper] [project]
Jan 22: Our work on Environment predictive coding is accepted at ICLR 2022. [paper] [project]
Oct 21: Our work on learning human activity priors for embodied AI is accepted at NeurIPS 2021. [paper]
Mar 21: Our work on bridging the egocentric <–> third-person gap in video is accepted at CVPR 2021. [paper]
Jan 21: Our work on Differentiable constraints for causal structure discovery is accepted at AISTATS 2021. [paper]
Sep 20: Our work on Exploration for Interaction is accepted at NeurIPS 2020. [paper] [project].
Jul 20: Invited talk at Stanford: “Environment-centric reasoning for embodied perception” [slides]. Thanks Juan Carlos Niebles for the invitation!
Mar 20: Our work on Topological environment graphs is accepted at CVPR 2020. [paper] [project]
Jul 19: Our work on Interaction Hotspots is accepted at ICCV 2019. [paper] [project]
Jun 19: Presented our work on Interaction Hotspots at the EPIC workshop at CVPR. EPIC@CVPR19
Jul 18: Our work on Attributes as Operators is accepted at ECCV 2018. [paper] [code]
May 18: Interning at Facebook AI Research this summer
Mar 18: Our work on Dynamic Inference Paths in ResNets is accepted at CVPR 2018. [paper] [code]
Dec 17: Our work on Computational Antimicrobial Peptide Design is accepted at JBC 2017. [paper] [code]
Aug 17: Completed a wonderful internship at IBM Research. Thanks Bhatta and Rogerio!
Jul 17: Sharmistha presents our work on Coupling Constraints for Relation Extraction at ACL 2017 (https://arxiv.org/abs/1710.09942)
May 17: Started an internship at IBM Research’s Vision & Multimedia group, working with Bishwaranjan Bhattacharjee and Dr. Rogerio Feris
Jan 17: Joined Dr. Kristen Grauman’s vision group as a Research Assistant.
Aug 16: Joined UT Austin as a graduate student.
July 16: Completed a wonderful internship at MALL Lab, thanks Partha!

Publications

	Step Differences in Instructional Video Tushar Nagarajan, Lorenzo Torresani CVPR 24 [paper] [data/code]
	Detours for Navigating Instructional Videos Kumar Ashutosh, Zihui Xue, Tushar Nagarajan, Kristen Grauman CVPR 24 (Highlight) [paper]
	Video ReCap: Recursive Captioning of Hour-Long Videos Md Mohaiminul Islam, Ngan Ho, Xitong Yang, Tushar Nagarajan, Lorenzo Torresani, Gedas Bertasius CVPR 24 [paper] [code]
	Ego-Exo4D: Understanding Skilled Human Activity from First-and Third-Person Perspectives Kristen Grauman, Andrew Westbury, Lorenzo Torresani, Kris Kitani, Jitendra Malik, Tushar Nagarajan* et al. CVPR 24 (Oral) [paper] [project]
	AnyMAL: An Efficient and Scalable Any-Modality Augmented Language Model Seungwhan Moon, Andrea Madotto, Zhaojiang Lin, Tushar Nagarajan, Matt Smith, Shashank Jain, Chun-Fu Yeh, Prakash Murugesan, Peyman Heidari, Yue Liu, Kavya Srinet, Babak Damavandi, Anuj Kumar ArXiv 23 (* equal contribution) [arxiv]
	HT-Step: Aligning Instructional Articles with How-To Videos Triantafyllos Afouras, Effrosyni Mavroudi, Tushar Nagarajan, Huiyu Wang, Lorenzo Torresani NeurIPS 23 [paper] [data/code]
	Ego4D Goal-Step: Toward Hierarchical Understanding of Procedural Activities Yale Song, Gene Byrne, Tushar Nagarajan, Huiyu Wang, Miguel Martin, Lorenzo Torresani NeurIPS 23 [paper] [code]
	EgoDistill: Egocentric Head Motion Distillation for Efficient Video Understanding Shuhan Tan, Tushar Nagarajan, Kristen Grauman NeurIPS 23 [paper] [project]
	EgoEnv: Human-centric environment representations from egocentric video Tushar Nagarajan, Santhosh K. Ramakrishnan, Ruta Desai, James Hillis, Kristen Grauman NeurIPS 23 (Oral) [paper] [project] [code]
	Ego4d: Around the world in 3,000 hours of egocentric video Kristen Grauman, Andrew Westbury, Tushar Nagarajan* , … Jitendra Malik CVPR 2022 (Oral) TPAMI 2023. Invited article: Best Papers of CVPR [paper] [project]
	Environment Predictive Coding for Visual Navigation Santhosh K. Ramakrishnan, Tushar Nagarajan, Ziad Al-Halah, Kristen Grauman ICLR 2022 [paper] [project] [code]
	Shaping embodied agent behavior with activity-context priors from egocentric video Tushar Nagarajan, Kristen Grauman NeurIPS 2021 (Spotlight) [paper] [project] [talk]
	Ego-Exo: Transferring Visual Representations from Third-person to First-person Videos Yanghao Li, Tushar Nagarajan, Bo Xiong, Kristen Grauman CVPR 2021 [paper] [code]
	Differentiable Causal Discovery Under Unmeasured Confounding Rohit Bhattacharya, Tushar Nagarajan, Daniel Malinsky, Ilya Shpitser AISTATS 2020 [paper] [code]
	Learning Affordance Landscapes for Interaction Exploration in 3D Environments Tushar Nagarajan, Kristen Grauman NeurIPS 2020 (Spotlight) [paper] [project] [talk] [code]
	Ego-Topo: Environment Affordances from Egocentric Video Tushar Nagarajan, Yanghao Li, Christoph Feichtenhofer, Kristen Grauman CVPR 2020 (Oral) [paper] [project] [talk] [code]
	Grounded Human-Object Interaction Hotspots from Video Tushar Nagarajan, Christoph Feichtenhofer, Kristen Grauman ICCV 2019 [paper] [project] [code]
	Attributes as Operators Tushar Nagarajan, Kristen Grauman ECCV 2018 [paper] [code]
	BlockDrop: Dynamic Inference Paths in Residual Networks Zuxuan Wu, Tushar Nagarajan, Abhishek Kumar, Steven Rennie, Larry S. Davis, Kristen Grauman, Rogerio Feris CVPR 2018 (Spotlight) (* equal contribution) [paper] [code] [talk]
	CANDiS: Coupled & Attention-Driven Neural Distant Supervision Tushar Nagarajan, Sharmistha Jat, Partha Talukdar ACL (Workshop) 2017 [paper]
	Computational antimicrobial peptide design and evaluation against multidrug-resistant clinical isolates of bacteria Deepesh Nagarajan, Tushar Nagarajan, Natasha Roy, Omkar Kulkarni, Sathyabaarathi Ravichandran, Madhulika Mishra, Dipshikha Chakravortty and Nagasuma Chandra Journal of Biological Chemistry (JBC), 2017. [paper] [code]