Zobrazeno 1 - 10
of 236
pro vyhledávání: '"Maji, Subhransu"'
Autor:
Hamilton, Max, Lange, Christian, Cole, Elijah, Shepard, Alexander, Heinrich, Samuel, Mac Aodha, Oisin, Van Horn, Grant, Maji, Subhransu
Species range maps (SRMs) are essential tools for research and policy-making in ecology, conservation, and environmental management. However, traditional SRMs rely on the availability of environmental covariates and high-quality species location obse
Externí odkaz:
http://arxiv.org/abs/2410.10931
We introduce One-Shot Label-Only (OSLO) membership inference attacks (MIAs), which accurately infer a given sample's membership in a target model's training set with high precision using just \emph{a single query}, where the target model only returns
Externí odkaz:
http://arxiv.org/abs/2405.16978
Autor:
Gregg, Benjamin, Calzetti, Daniela, Adamo, Angela, Bajaj, Varun, Ryon, Jenna E., Linden, Sean T., Correnti, Matteo, Cignoni, Michele, Messa, Matteo, Sabbi, Elena, Gallagher, John S., Grasha, Kathryn, Pedrini, Alex, Gutermuth, Robert A., Melinder, Jens, Kotulla, Ralf, Pérez, Gustavo, Krumholz, Mark R., Bik, Arjan, Östlin, Göran, Johnson, Kelsey E., Bortolini, Giacomo, Smith, Linda J., Tosi, Monica, Maji, Subhransu, Vieira, Helena Faustino
We present maps of ionized gas (traced by Pa$\alpha$ and Br$\alpha$) and 3.3 $\mu$m Polycyclic Aromatic Hydrocarbon (PAH) emission in the nearby spiral galaxy NGC 628, derived from new JWST/NIRCam data from the FEAST survey. With this data, we invest
Externí odkaz:
http://arxiv.org/abs/2405.09667
Modeling and visualizing relationships between tasks or datasets is an important step towards solving various meta-tasks such as dataset discovery, multi-tasking, and transfer learning. However, many relationships, such as containment and transferabi
Externí odkaz:
http://arxiv.org/abs/2403.17173
The zero-shot performance of existing vision-language models (VLMs) such as CLIP is limited by the availability of large-scale, aligned image and text datasets in specific domains. In this work, we leverage two complementary sources of information --
Externí odkaz:
http://arxiv.org/abs/2401.02460
Computer vision-based re-identification (Re-ID) systems are increasingly being deployed for estimating population size in large image collections. However, the estimated size can be significantly inaccurate when the task is challenging or when deploy
Externí odkaz:
http://arxiv.org/abs/2312.05287
Autor:
Saha, Oindrila, Maji, Subhransu
We develop techniques for refining representations for fine-grained classification and segmentation tasks in a self-supervised manner. We find that fine-tuning methods based on instance-discriminative contrastive learning are not as effective, and po
Externí odkaz:
http://arxiv.org/abs/2309.13822
We present a set of metrics that utilize vision priors to effectively assess the performance of saliency methods on image classification tasks. To understand behavior in deep learning models, many methods provide visual saliency maps emphasizing imag
Externí odkaz:
http://arxiv.org/abs/2309.10989
Autor:
Cheng, Zezhou, Esteves, Carlos, Jampani, Varun, Kar, Abhishek, Maji, Subhransu, Makadia, Ameesh
A critical obstacle preventing NeRF models from being deployed broadly in the wild is their reliance on accurate camera poses. Consequently, there is growing interest in extending NeRF models to jointly optimize camera poses and scene representation,
Externí odkaz:
http://arxiv.org/abs/2306.05410
Many modern applications use computer vision to detect and count objects in massive image collections. However, when the detection task is very difficult or in the presence of domain shifts, the counts may be inaccurate even with significant investme
Externí odkaz:
http://arxiv.org/abs/2306.03151