Subhankar Roy

University of Bergamo — Italy

Computer Science · Physical Sciences

13h-index1.3kcitations53works3.12yr avg
OpenAlex↗ORCID↗Google Scholar↗Lab page↗

Accepting Students?

No reports yet. Know if this professor is taking students?

Research Topics

Domain Adaptation and Few-Shot Learning(37), Multimodal Machine Learning Applications(22), Cancer-related molecular mechanisms research(8), Human Pose and Action Recognition(6), Advanced Neural Network Applications(5)

Publications53 total

LT-Soups: Bridging Head and Tail Classes via Subsampled Model Soups
arXiv (Cornell University)·2025Open Access
Ensembling Pruned Attention Heads For Uncertainty-Aware Efficient Transformers
arXiv (Cornell University)·2025Open Access
Make me an Expert: Distilling from Generalist Black-Box Models into Specialized Models for Semantic Segmentation
ArXiv.org·2025Open Access
FedMVP: Federated Multimodal Visual Prompt Tuning for Vision-Language Models
arXiv (Cornell University)·2025Open Access
Group-robust Machine Unlearning
arXiv (Cornell University)·2025Open Access
Uncertainty quantification in deep neural networks for multi-sensor Earth observation
Elsevier eBooks·2025· 2 cited
Large-Scale Pre-trained Models are Surprisingly Strong in Incremental Novel Class Discovery
Lecture notes in computer science·2024· 5 cited
Weighted Ensemble Models Are Strong Continual Learners
Lecture notes in computer science·2024· 9 cited
Organizing Unstructured Image Collections using Natural Language
arXiv (Cornell University)·2024· 1 citedOpen Access
Unlearning Personal Data from a Single Image
arXiv (Cornell University)·2024Open Access
Collaborating Foundation Models for Domain Generalized Semantic Segmentation
2024· 17 citedOpen Access
Less is more: Summarizing Patch Tokens for efficient Multi-Label Class-Incremental Learning
arXiv (Cornell University)·2024· 1 citedOpen Access
Simplifying open-set video domain adaptation with contrastive learning
Computer Vision and Image Understanding·2024· 8 citedOpen Access
Democratizing Fine-grained Visual Recognition with Large Language Models
arXiv (Cornell University)·2024· 1 citedOpen Access
Collaborating Foundation Models for Domain Generalized Semantic Segmentation
arXiv (Cornell University)·2023Open Access
Weighted Ensemble Models Are Strong Continual Learners
arXiv (Cornell University)·2023Open Access
Enhancing Plasticity for First Session Adaptation Continual Learning
arXiv (Cornell University)·2023Open Access
The Unreasonable Effectiveness of Large Language-Vision Models for Source-free Video Domain Adaptation
2023· 13 citedOpen Access
The Unreasonable Effectiveness of Large Language-Vision Models for Source-free Video Domain Adaptation
arXiv (Cornell University)·2023· 2 citedOpen Access
Contrast, Stylize and Adapt: Unsupervised Contrastive Learning Framework for Domain Adaptive Semantic Segmentation
arXiv (Cornell University)·2023Open Access
page 1 of 3Next →

Frequent Co-authors

Elisa Ricci(33), Nicu Sebe(22), Stéphane Lathuilière(16), Paolo Rota(7), Enver Sangineto(7), Zhun Zhong(7), Mingxuan Liu(7), Aliaksandr Siarohin(5), Yasser Benigmim(5), Slim Essid(5), Vicky Kalogeiton(5), Giacomo Zara(5), Enrico Fini(4), Begüm Demir(4), Andrea Pilzer(4), Zhun Zhong(4), Hongtao Lu(4), Martin Trapp(3), Arno Solin(3), Thomas De Min(3)