Sagnik Majumder

I am a PhD student in Computer Science at UT Austin, advised by Prof. Kristen Grauman. Before this, I received my MS in Computer Science at UT. I am broadly interested in computer vision and machine learning. My current line of research is embodied audio-visual understanding of 3D scenes with applications in mobile robotics and AR/VR.

Previously, I have worked with Prof. Visvanathan Ramesh at Goethe University on continual and meta learning for image recognition tasks. I have also had the pleasure of collaborating with Prof. Christoph Malsburg at the Frankfurt Institute for Advanced Studies for investigating visual models that draw motivation from Neuroscience.

Earlier, I graduated from BITS Pilani.

Industrial research internships: I am looking for industrial research internships in audio-visual/multi-modal learning for Summer 2025. Kindly reach out if you think I would be a good fit.

Research collaborations: I am open to collaborating on research projects, and also mentoring Master's and final (senior) year students on their thesis. Shoot me an email to discuss more.

CV | E-Mail | Google Scholar | Github | Twitter

Affiliations


BITS Pilani 2014-2018	FIAS Summer 2017	Goethe University 2018-2019	Meta AI 2022-present	UT Austin 2019-present

News

Mar 2024	Our paper Learning Spatial Features from Audio-Visual Correspondence in Egocentric Videos got accepted at CVPR 24!
Mar 2024	Our paper Ego-Exo4D got accepted at CVPR 24!
Mar 2023	Our paper Chat2Map: Efficient Scene Mapping from Multi-Ego Conversations got accepted at CVPR 23!
Feb 2023	Co-organzing the SoundSpaces Challenge at the CVPR 2023 Embodied AI Workshop.
June 2023	Invited talk at CVPR 23 Sight and Sound Workshop, "Chat2Map: Efficient Scene Mapping from Multi-Ego Conversations".
May 2023	Invited talk at JHU NSA Lab, "Efficiently understanding 3D scenes using sight and sound".
Mar 2023	Our paper Chat2Map: Efficient Scene Mapping from Multi-Ego Conversations. got accepted at CVPR 23!
Feb 2023	Co-organzing the CVPR 2023 SoundSpaces Challenge and the Embodied AI Workshop.
Dec 2022	Starting as a visiting researcher at Meta AI Research.
Oct 2022	Invited talk at ECCV 22 AV4D Workshop, "Active Audio-Visual Separation of Dynamic Sound Sources".
Sept 2022	Our paper Few-Shot Audio-Visual Learning of Environment Acoustics got accepted at NeurIPS 22!
Sept 2022	Continuing as a student researcher at Meta Reality Labs Redmond this Fall.
July 2022	Our paper Active Audio-Visual Separation of Dynamic Sound Sources got accepted at ECCV 22!
June 2022	Invited talk at CVPR 22 Sight and Sound Workshop, "Active Audio-Visual Separation of Dynamic Sound Sources" (Slides).
June 2022	Joined Meta Reality Labs Redmond as a research scientist intern this summer.
April 2022	Invited talk at Meta AI Research, "Active Audio-Visual Separation of Dynamic Sound Sources" (Slides).
Feb 2022	Co-organzing the SoundSpaces Challenge at the CVPR 2022 Embodied AI Workshop.

Publications

	Learning Spatial Features from Audio-Visual Correspondence in Egocentric Videos Sagnik Majumder, Ziad Al-Halah, Kristen Grauman CVPR 2024 paper \| project \| code and data (coming soon!)
	Ego-Exo4D: Understanding Skilled Human Activity from First- and Third-Person Perspectives Kristen Grauman, Andrew Westbury, Lorenzo Torresani, ..., Sagnik Majumder, ..., Mike Zheng Shou, Michael Wray CVPR 2024 paper \| project
	Chat2Map: Efficient Scene Mapping from Multi-Ego Conversations Sagnik Majumder, Hao Jiang, Pierre Moulon, Ethan Henderson, Paul Calamia, Kristen Grauman, Vamsi Krishna Ithapu Equal contribution CVPR 2023* paper \| project \| code and data
	Retrospectives on the Embodied AI Workshop Matt Deitke, Dhruv Batra, Yonatan Bisk, ..., Sagnik Majumder, ..., Luca Weihs, Jiajun Wu arXiv paper
	Few-Shot Audio-Visual Learning of Environment Acoustics Sagnik Majumder, Changan Chen, Ziad Al-Halah, Kristen Grauman Equal contribution NeurIPS 2022* paper \| project \| code and data
	Active Audio-Visual Separation of Dynamic Sound Sources Sagnik Majumder, Kristen Grauman ECCV 2022 paper \| project \| code and data
	Move2Hear: Active Audio-Visual Source Separation Sagnik Majumder, Ziad Al-Halah, Kristen Grauman ICCV 2021 paper \| project \| code and data
	Learning to Set Waypoints for Audio-Visual Navigation Changan Chen, Sagnik Majumder, Ziad Al-Halah, Ruohan Gao, Santhosh K. Ramakrishnan, Kristen Grauman ICLR 2021 paper \| project \| code
	Model Agnostic Answer Reranking System for Adversarial Question Answering Sagnik Majumder, Chinmoy Samant, Greg Durrett EACL SRW 2021 paper
	Meta-learning Convolutional Neural Architectures for Multi-target Concrete Defect Classification with the COncrete DEfect BRidge IMage Dataset Martin Mundt, Sagnik Majumder, Sreenivas Murali, Panagiotis Panetsos, Visvanathan Ramesh CVPR 2021 paper \| code and data
	Unified Probabilistic Deep Continual Learning through Generative Replay and Open Set Recognition Martin Mundt, Sagnik Majumder, Iuliia Pliushch, Visvanathan Ramesh arXiv 2019 paper \| code
	Open Set Recognition Through Deep Neural Network Uncertainty: Does Out-of-Distribution Detection Require Generative Classifiers? Martin Mundt, Iuliia Pliushch, Sagnik Majumder, Visvanathan Ramesh ICCV SDLCV Workshop 2019 paper \| code
	Rethinking Layer-wise Feature Amounts in Convolutional Neural Network Architectures Martin Mundt, Sagnik Majumder, Tobias Weis, Visvanathan Ramesh NeurIPS CRACT Workshop 2018 paper \| code
	Handwritten Digit Recognition by Elastic Matching Sagnik Majumder, Christoph von der Malsburg, Aashish Richhariya, Surekha Bhanot JCP 2018 paper \| code

Template credits: Unnat, Changan and Jon