Skip to content
VirtualOnlinePast event

Advances in AI at Johns Hopkins University

Event artwork for Advances in AI at Johns Hopkins University

This event has ended

This listing is kept as an archive. Explore upcoming events to find your next one.

See upcoming events
About this event

What to expect

Explore cutting-edge AI breakthroughs from Johns Hopkins University researchers in this comprehensive virtual meetup covering state-of-the-art advances in image generation, model personalization, and trustworthy AI systems. This online event brings together experts presenting their latest research on diffusion models, representation learning, computational pathology, and real-world data challenges.

Featured Research Presentations

Recent Advancements in Image Generation and Understanding — Aniket Roy, Research Scientist at NEC Labs America and recent PhD graduate from Johns Hopkins, presents three groundbreaking works. Learn about DiffNat (TMLR'25), which leverages kurtosis concentration properties to enhance diffusion models for sharper, more faithful image generation. Discover DuoLoRA (ICCV'25), an adaptive-rank approach to content-style personalization that achieves superior results with 19× fewer parameters. Explore Cap2Aug (WACV'25), a novel semantic augmentation technique using caption-based backtranslation to address data scarcity in few-shot and long-tail classification scenarios.

From Representation Analysis to Data Refinement — Basudha Pal, PhD from Johns Hopkins ECE, examines how deep learning models encode information beyond their training objectives. This presentation introduces mutual information-based expressivity estimation to quantify implicit attribute capture in learned embeddings, revealing hierarchical patterns across network layers. Learn data attribution techniques for identifying influential training samples that contribute to model errors, with practical applications for improving robustness, fairness, and reliability in biometric and medical imaging systems.

Scalable & Precise Histopathology — Tawsifur Rahman, PhD candidate in Biomedical Engineering at Johns Hopkins, addresses the computational challenges of gigapixel whole slide images in digital pathology. Discover the Task-Specific Self-Supervised Learning (TS-SSL) framework that uses spatial-channel attention for domain-optimized feature representations, outperforming existing foundation models. Explore CEMIL, a contextual attention-based Multiple Instance Learning framework using instructor-learner knowledge distillation to achieve state-of-the-art cancer classification with significantly reduced computa

Running an AI event?

List it here, free.

Meetups, workshops and conferences go in front of thousands of builders after a quick review.

List your event