Explanipedia

Merlin L48 Spectrogram Dataset Open

Aaron Sun, Subhransu Maji, Grant Van Horn · 2025

In the single-positive multi-label (SPML) setting, each image in a dataset is labeled with the presence of a single class, while the true presence of other classes remains unknown. The challenge is to narrow the performance gap between thi…

Pairing weather radars with eBird data for large-scale monitoring of swallow and martin communal roosts Open

Maria Carolina T. D. Belotti, Wenlong Zhao, Yuting Deng, Victoria F. Simons, Gustavo Pérez , et al. · 2025

During their nonbreeding period, many species of swallows and martins (family: Hirundinidae) congregate in large communal roosts. Some of these roosts are well-known within local birdwatching communities; however, monitoring them at large …

SIGMA-GEN: Structure and Identity Guided Multi-subject Assembly for Image Generation Open

Oindrila Saha, Vojtech Krs, Radomír Měch, Subhransu Maji, Kevin Blackburn-Matzen , et al. · 2025

We present SIGMA-GEN, a unified framework for multi-identity preserving image generation. Unlike prior approaches, SIGMA-GEN is the first to enable single-pass multi-subject identity-preserved generation guided by both structural and spati…

Reinforcement Learning for Quantum Network Control with Application-Driven Objectives Open

Guo Xian Yau, Alexandra Burushkina, Francisco Ferreira da Silva, Subhransu Maji, Philip S. Thomas , et al. · 2025

Optimized control of quantum networks is essential for enabling distributed quantum applications with strict performance requirements. In near-term architectures with constrained hardware, effective control may determine the feasibility of…

Consensus-Driven Active Model Selection Open

Justin Kay, Grant Van Horn, Subhransu Maji, Daniel Sheldon, Sara Beery · 2025

The widespread availability of off-the-shelf machine learning models poses a challenge: which model, of the many available candidates, should be chosen for a given data analysis task? This question of model selection is traditionally answe…

The iNaturalist Sounds Dataset Open

Mustafa Chasmai, Alexander Shepard, Subhransu Maji, Grant Van Horn · 2025

We present the iNaturalist Sounds Dataset (iNatSounds), a collection of 230,000 audio files capturing sounds from over 5,500 species, contributed by more than 27,000 recordists worldwide. The dataset encompasses sounds from birds, mammals,…

Representation Learning for Long-Chain Hydrocarbon Adsorption in Zeolites Open

Yachan Liu, Ping Yang, Gustavo Pérez, Aaron Sun, Wei Fan , et al. · 2025

Zeolites are a class of crystalline nanoporous materials known for their ability to discriminate molecules based on size and shape. Such molecular shape selectivity arises from the precise 3-dimensional arrangement of zeolite framework ato…

Audio Geolocation: A Natural Sounds Benchmark Open

Mustafa Chasmai, W.-Y. Liu, Subhransu Maji, Grant Van Horn · 2025

Can we determine someone's geographic location purely from the sounds they hear? Are acoustic signals enough to localize within a country, state, or even city? We tackle the challenge of global-scale audio geolocation, formalize the proble…

Automated extraction of right whale morphometric data from drone aerial photographs Open

Chhandak Bagchi, Josh Medina, Duncan J. Irschick, Subhransu Maji, Fredrik Christiansen · 2025

Aerial photogrammetry is a popular non‐invasive tool to measure the size, body morphometrics and body condition of wild animals. While the method can generate large datasets quickly, the lack of efficient processing tools can create bottle…

CVOCSemRPL: Class-Variance Optimized Clustering, Semantic Information Injection and Restricted Pseudo Labeling based Improved Semi-Supervised Few-Shot Learning Open

R. K. Baghel, Subhransu Maji, Pratik Mazumder · 2025

Few-shot learning has been extensively explored to address problems where the amount of labeled samples is very limited for some classes. In the semi-supervised few-shot learning setting, substantial quantities of unlabeled samples are ava…

Improving Satellite Imagery Masking Using Multitask and Transfer Learning Open

Rangel Daroya, Luísa Vieira Lucchese, Travis Simmons, Punwath Prum, Tamlin M. Pavelsky , et al. · 2025

Many remote sensing applications require masking of pixels in satellite imagery for further analysis. For instance, estimating water quality variables such as suspended sediment concentration (SSC) requires isolating pixels depicting water…

WildSAT: Learning Satellite Image Representations from Wildlife Observations Open

Rangel Daroya, Elijah Cole, Oisin Mac Aodha, Grant Van Horn, Subhransu Maji · 2024

Species distributions encode valuable ecological and environmental information, yet their potential for guiding representation learning in remote sensing remains underexplored. We introduce WildSAT, which pairs satellite images with millio…

Improving Satellite Imagery Masking using Multi-task and Transfer Learning Open

Rangel Daroya, Luísa Vieira Lucchese, Travis Simmons, Punwath Prum, Tamlin M. Pavelsky , et al. · 2024

Many remote sensing applications employ masking of pixels in satellite imagery for subsequent measurements. For example, estimating water quality variables, such as Suspended Sediment Concentration (SSC) requires isolating pixels depicting…

Combining Observational Data and Language for Species Range Estimation Open

Max Hamilton, Christian Lange, Elijah Cole, Alexander D. Shepard, S. Heinrich , et al. · 2024

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 lo…

OSLO: One-Shot Label-Only Membership Inference Attacks Open

Yuefeng Peng, Jaechul Roh, Subhransu Maji, Amir Houmansadr · 2024

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 o…

Feedback in emerging extragalactic star clusters, FEAST: The relation between 3.3 $μ$m PAH emission and Star Formation Rate traced by ionized gas in NGC 628 Open

Benjamin Gregg, Daniela Calzetti, Angela Adamo, Varun Bajaj, J. E. Ryon , et al. · 2024

We present maps of ionized gas (traced by Pa$α$ and Br$α$) and 3.3 $μ$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 investi…

Task2Box: Box Embeddings for Modeling Asymmetric Task Relationships Open

Rangel Daroya, Aaron Sun, Subhransu Maji · 2024

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 …

DISCount: Counting in Large Image Collections with Detector-Based Importance Sampling Open

Gustavo Pérez, Subhransu Maji, Daniel Sheldon · 2024

Many applications use computer vision to detect and count objects in massive image collections. However, automated methods may fail to deliver accurate counts, especially when the task is very difficult or requires a fast response time. Fo…

Modeling suspended sediment concentration using artificial neural networks, an effort towards global sediment flux observations in rivers from space Open

Luísa Vieira Lucchese, Rangel Daroya, Travis Simmons, Punwath Prum, Subhransu Maji , et al. · 2024

Harmonized Landsat Sentinel-2 (HLS) provides high-quality images every 2-3 days across Earth. However, HLS has not been widely used to measure Suspended Sediment Concentration (SSC) in rivers. Here, we used HLS to generate a fully open-sou…

Improved Zero-Shot Classification by Adapting VLMs with Text Descriptions Open

Oindrila Saha, Grant Van Horn, Subhransu Maji · 2024

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 inf…

Human-in-the-Loop Visual Re-ID for Population Size Estimation Open

Gustavo Pérez, Daniel Sheldon, Grant Van Horn, Subhransu Maji · 2023

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 …

PARTICLE: Part Discovery and Contrastive Learning for Fine-grained Recognition Open

Oindrila Saha, Subhransu Maji · 2023

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 effect…

COSE: A Consistency-Sensitivity Metric for Saliency on Image Classification Open

Rangel Daroya, Aaron Sun, Subhransu Maji · 2023

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 empha…

LU-NeRF: Scene and Pose Estimation by Synchronizing Local Unposed NeRFs Open

Zezhou Cheng, Carlos Esteves, Varun Jampani, Abhishek Kar, Subhransu Maji , et al. · 2023

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 repr…

DISCount: Counting in Large Image Collections with Detector-Based Importance Sampling Open

Gustavo Pérez, Subhransu Maji, Daniel Sheldon · 2023

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 significa…

Inside back cover Open

P. Bai, Subhransu Maji · 2023

Permissions Request permissions Inside back cover J. Mater. Chem. A, 2023, 11, 17873 DOI: 10.1039/D3TA90179C This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. You can use material from this article in othe…

Accidental Turntables: Learning 3D Pose by Watching Objects Turn Open

Zezhou Cheng, Matheus Gadelha, Subhransu Maji · 2022

We propose a technique for learning single-view 3D object pose estimation models by utilizing a new source of data -- in-the-wild videos where objects turn. Such videos are prevalent in practice (e.g., cars in roundabouts, airplanes near r…

Using spatio-temporal information in weather radar data to detect and track communal bird roosts Open

Gustavo Pérez, Wenlong Zhao, Zezhou Cheng, Maria Carolina T. D. Belotti, Yuting Deng , et al. · 2022

The exodus of swallows from communal nighttime roosts is often visible as an expanding ring-shaped pattern in weather radar data. The WSR-88D network operated by the National Weather Service archives more than 25 years of data across 143 s…

MvDeCor: Multi-view Dense Correspondence Learning for Fine-grained 3D Segmentation Open

G. K. Sharma, Kangxue Yin, Subhransu Maji, Evangelos Kalogerakis, Or Litany , et al. · 2022

We propose to utilize self-supervised techniques in the 2D domain for fine-grained 3D shape segmentation tasks. This is inspired by the observation that view-based surface representations are more effective at modeling high-resolution surf…

Star Cluster Formation and Evolution in M101: An Investigation with the Legacy Extragalactic UV Survey Open

Sean T. Linden, Gustavo Pérez, Daniela Calzetti, Subhransu Maji, Matteo Messa , et al. · 2022

We present Hubble Space Telescope WFC3/UVIS (F275W, F336W) and ACS/WFC optical (F435W, F555W, and F814W) observations of the nearby grand-design spiral galaxy M101 as part of the Legacy Extragalactic UV Survey (LEGUS). Compact sources dete…

Subhransu Maji YOU? Author Swipe