Anubhav Gupta
Research Subjects
Anubhav Gupta has a total of 25 co-authors. Here are the top co-authors by number of shared publications:
- Shishira R. Maiya
University of Maryland University of Maryland University of Maryland University of Maryland
3 shared publications - Abhinav Shrivastava
University of Maryland College Park
2 shared publications - Archana Swaminathan
University of Maryland
2 shared publications - Kamal Gupta
University of Maryland
2 shared publications - Vatsal Agarwal
University of Maryland
2 shared publications Abhinav Shrivastava
University of Maryland 20742 College Park MD USA
1 shared publication
Latest Publications
- 22 November 2024
- Published by Springer Nature
- p. 285-302
Latent-INR: A Flexible Framework for Implicit Representations of Videos with Discriminative Semantics
- 22 November 2024
- Published by Springer Nature
- p. 285-302
Latent-INR: A Flexible Framework for Implicit Representations of Videos with Discriminative Semantics
- 29 October 2024
- Published by Springer Nature
- p. 210-227
LEIA: Latent View-Invariant Embeddings for Implicit 3D Articulation
- 29 October 2024
- Published by Springer Nature
- p. 210-227
LEIA: Latent View-Invariant Embeddings for Implicit 3D Articulation
- Open Access
LEIA: Latent View-invariant Embeddings for Implicit 3D Articulation
Neural Radiance Fields (NeRFs) have revolutionized the reconstruction of static scenes and objects in 3D, offering unprecedented quality. However, extending NeRFs to model dynamic objects or object articulations remains a challenging problem. Previous works have tackled this issue by focusing on part-level reconstruction and motion estimation for objects, but they often rely on heuristics regarding the number of moving parts or object categories, which can limit their practical use. In this work, we introduce LEIA, a novel approach for representing dynamic 3D objects. Our method involves observing the object at distinct time steps or "states" and conditioning a hypernetwork on the current state, using this to parameterize our NeRF. This approach allows us to learn a view-invariant latent representation for each state. We further demonstrate that by interpolating between these states, we can generate novel articulation configurations in 3D space that were previously unseen. Our experimental results highlight the effectiveness of our method in articulating objects in a manner that is independent of the viewing angle and joint configuration. Notably, our approach outperforms previous methods that rely on motion information for articulation registration.
Open AccessLEIA: Latent View-invariant Embeddings for Implicit 3D Articulation
Neural Radiance Fields (NeRFs) have revolutionized the reconstruction of static scenes and objects in 3D, offering unprecedented quality. However, extending NeRFs to model dynamic objects or object articulations remains a challenging problem. Previous works have tackled this issue by focusing on part-level reconstruction and motion estimation for objects, but they often rely on heuristics regarding the number of moving parts or object categories, which can limit their practical use. In this work, we introduce LEIA, a novel approach for representing dynamic 3D objects. Our method involves observing the object at distinct time steps or "states" and conditioning a hypernetwork on the current state, using this to parameterize our NeRF. This approach allows us to learn a view-invariant latent representation for each state. We further demonstrate that by interpolating between these states, we can generate novel articulation configurations in 3D space that were previously unseen. Our experimental results highlight the effectiveness of our method in articulating objects in a manner that is independent of the viewing angle and joint configuration. Notably, our approach outperforms previous methods that rely on motion information for articulation registration.
- Open Access
- 12 April 2023
- Published by The Open Journal in Journal of Open Source Software
- Vol. 8 (84) , 4912
icepyx: querying, obtaining, analyzing, and manipulating ICESat-2 datasets
Icepyx is both a software library and a community composed of ICESat-2 (NASA satellite) data users, developers, maintainers, and the scientific community. We are working together to develop a shared library of resources-including existing resources, new code, tutorials, and use-cases/examples-that simplify the process of querying, obtaining, analyzing, and manipulating ICESat-2 datasets to enable scientific discovery.
- 12 April 2023
- Published by The Open Journal in Journal of Open Source Software
- Vol. 8 (84) , 4912
Open Accessicepyx: querying, obtaining, analyzing, and manipulating ICESat-2 datasets
Icepyx is both a software library and a community composed of ICESat-2 (NASA satellite) data users, developers, maintainers, and the scientific community. We are working together to develop a shared library of resources-including existing resources, new code, tutorials, and use-cases/examples-that simplify the process of querying, obtaining, analyzing, and manipulating ICESat-2 datasets to enable scientific discovery.