We value your privacy

We use cookies to understand how you interact with our website to improve your experience. By accepting, you agree to our use of these cookies. You can always change your mind later.

IFLAI Research & Technical Insights

Engineering Intelligence

Deep dives into data-efficient foundation models, edge deployment strategies, and the future of computer vision without massive datasets.

The Case for AI That Learns More From Less

InsightsMay 26, 2026

The Case for AI That Learns More From Less

Modern AI has been shaped by scale: more data, more compute, more parameters. But in science and industry, the next leap may come from models that are smaller, more efficient, and better grounded in the structure of the problem.

IFLAIRead Article

The AI Risk Framing Problem: Extremes and Agendas

InsightsMay 12, 2026

The AI Risk Framing Problem: Extremes and Agendas

Discussions around AI risk are often framed in apocalyptic or utopian extremes. A closer look at the data reveals why neither narrative reflects reality.

IFLAIRead Article

Why 5 Datapoints Outperforms 50,000: Physics-Informed Priors, Not Fine-Tuning

IFLAI ResearchApril 24, 2026

Why 5 Datapoints Outperforms 50,000: Physics-Informed Priors, Not Fine-Tuning

IFLAI's approach to data-efficient microscopy AI isn't foundation model fine-tuning. It's something fundamentally different: building physics into the architecture itself.

IFLAI ResearchRead Article

Label-Free Characterization of Few-kDa Biomolecules via Hierarchical Vision Transformers

IFLAI ResearchMarch 15, 2026

Label-Free Characterization of Few-kDa Biomolecules via Hierarchical Vision Transformers

A deep dive into our recent Nature Communications paper demonstrating how Vision Transformers augment nanofluidic scattering microscopy to achieve unprecedented mass and size characterization.

IFLAI ResearchRead Article

How Physics-Informed AI Accelerates Drug Discovery at AstraZeneca

IFLAI ResearchFebruary 9, 2026

How Physics-Informed AI Accelerates Drug Discovery at AstraZeneca

A look at our recent feature in Invest in Gothenburg, detailing our collaboration with AstraZeneca and how integrating physical priors into AI architectures dramatically reduces the data required for automated drug screening.

Henrik Klein Moberg (CTO)Read Article

Securing the Future: Detecting Zero-Day Physical Threats with AI

IFLAI ResearchJanuary 20, 2026

Securing the Future: Detecting Zero-Day Physical Threats with AI

How active learning and edge-deployed computer vision are enabling security teams to respond to novel threats without months of retraining.

IFLAI Security LabsRead Article

Advancing Science with Data-Efficient AI: CHAIR 2025 Keynote

EventsDecember 19, 2025

Advancing Science with Data-Efficient AI: CHAIR 2025 Keynote

Watch our CTO, Henrik Klein Moberg, present our core philosophy on physics-informed architectures at the Chalmers AI Research Centre (CHAIR) Students of AI 2025 event.

IFLAIRead Article

Inductive Biases for Efficient Deep Learning in Microscopy

IFLAI ResearchJune 10, 2025

Inductive Biases for Efficient Deep Learning in Microscopy

A deep dive into how baking physical intuition into neural network architectures allows for extreme data efficiency in microscopy.

IFLAI ResearchRead Article

Spatial Clustering of Single-Molecule Localizations via Graph Neural Networks

IFLAI ResearchMay 22, 2025

Spatial Clustering of Single-Molecule Localizations via Graph Neural Networks

Executive Summary: Deploying Graph Neural Networks to out-perform classical density-based clustering in Single-Molecule Localization Microscopy.

IFLAI ResearchRead Article

Analyzing Motion Changes in Single-Particle Experiments

IFLAI ResearchApril 10, 2025

Analyzing Motion Changes in Single-Particle Experiments

The story behind our Nature Communications paper on why detecting sudden state changes in diffusing nanoparticles is so computationally demanding.

IFLAI ResearchRead Article

Deep Dive into AI: Equipping Chalmers Faculty for the AI Era

NewsMarch 15, 2025

Deep Dive into AI: Equipping Chalmers Faculty for the AI Era

A successful kick-off to the 'Deep Dive into AI' course at Chalmers University of Technology, an initiative designed to integrate deep learning across scientific research domains.

IFLAIRead Article

Accelerating Plasmonic Hydrogen Sensors via Transformer-Based Deep Learning

IFLAI ResearchMarch 1, 2025

Accelerating Plasmonic Hydrogen Sensors via Transformer-Based Deep Learning

A technical breakdown of how we applied Vision Transformers to temporal optical data to vastly accelerate hydrogen leak detection.

IFLAI ResearchRead Article

Online Mass Spectrometry of Single Catalyst Nanoparticles via Deep Learning

IFLAI ResearchFebruary 15, 2025

Online Mass Spectrometry of Single Catalyst Nanoparticles via Deep Learning

A case study on achieving real-time mass spectrometry at the single-nanoparticle level using specialized neural denoising.

IFLAI ResearchRead Article

Sustainability Circle: Great AI Doesn't Need Great Data

EventsFebruary 6, 2025

Sustainability Circle: Great AI Doesn't Need Great Data

A presentation by Mattias Goksör and Henrik Klein Moberg on how data-efficient AI is paving the way for a sustainable, cost-effective industrial future.

IFLAIRead Article

Demystifying AI: The Deep Learning Crash Course

NewsJanuary 10, 2025

Demystifying AI: The Deep Learning Crash Course

A note from the authors on why we wrote a hands-on introduction to modern deep learning for the scientific community.

IFLAIRead Article

Pushing the Boundaries of Single Molecule Microscopy through Deep Learning

IFLAI ResearchNovember 15, 2024

Pushing the Boundaries of Single Molecule Microscopy through Deep Learning

Reviewing the 2024 PhD Thesis that unified deep learning approaches across single-molecule microscopy, optical sensing, and mass spectrometry.

IFLAI ResearchRead Article

Materials for Tomorrow 2024: The Future of AI in Material Science

EventsOctober 31, 2024

Materials for Tomorrow 2024: The Future of AI in Material Science

Insights from our presentation at the Materials for Tomorrow Conference at Chalmers University of Technology, exploring the intersection of deep learning and advanced material analysis.

IFLAIRead Article

IFLAI at Gothenburg Tech Week 2024

EventsOctober 4, 2024

IFLAI at Gothenburg Tech Week 2024

Connecting with the ecosystem and sharing our vision for sustainable, data-efficient AI at one of Europe's most vibrant tech events.

IFLAIRead Article

Annotation-Free Deep Learning for Quantitative Microscopy

IFLAI ResearchSeptember 5, 2024

Annotation-Free Deep Learning for Quantitative Microscopy

Exploring the 2024 PhD Thesis that formalized the transition from human-annotated datasets to self-supervised learning in computational biology.

IFLAI ResearchRead Article

Cross-Modality Transformations in Biological Microscopy

IFLAI ResearchJune 20, 2024

Cross-Modality Transformations in Biological Microscopy

How generative deep learning models allow researchers to synthetically translate brightfield images into fluorescent modalities, saving time and reagents.

IFLAI ResearchRead Article

Achieving 10X Sensitivity in Nanoplasmonic Sensors

IFLAI ResearchFebruary 15, 2024

Achieving 10X Sensitivity in Nanoplasmonic Sensors

How neural networks drastically amplify the sensitivity of optical hydrogen sensors, proving that software can outscale hardware improvements.

IFLAI ResearchRead Article

Geometric Deep Learning Reveals Spatiotemporal Features of Motion

IFLAI ResearchOctober 15, 2023

Geometric Deep Learning Reveals Spatiotemporal Features of Motion

How Geometric Deep Learning unlocks a fundamentally new understanding of anomalous diffusion and microscopic motion in complex environments.

IFLAI ResearchRead Article

Eliminating the Annotation Bottleneck: Single-Shot Self-Supervised Object Detection

IFLAI ResearchNovember 20, 2022

Eliminating the Annotation Bottleneck: Single-Shot Self-Supervised Object Detection

An inside look into our 2022 Nature Communications paper, formatted as a Q&A on why we abandoned human annotation in microscopy.

IFLAI ResearchRead Article

IFLAI at ETH Zurich: Advanced Machine Learning for Microscopy

NewsOctober 10, 2022

IFLAI at ETH Zurich: Advanced Machine Learning for Microscopy

Teaching the next generation of researchers how to apply deep learning to microscopy data at the MaP Doctoral School, ETH Zurich.

IFLAIRead Article

Label-Free Nanofluidic Scattering Microscopy

IFLAI ResearchApril 10, 2022

Label-Free Nanofluidic Scattering Microscopy

Our 2022 Nature Methods paper that demonstrated how optical scattering can be used to measure the mass and size of single diffusing molecules without labels.

IFLAI ResearchRead Article

Extracting Quantitative Data from Bright-Field Cells via Deep Learning

IFLAI ResearchAugust 5, 2021

Extracting Quantitative Data from Bright-Field Cells via Deep Learning

Reviewing our 2021 Biophysical Reviews paper on transforming qualitative bright-field microscopy into strict, quantitative biological data.

IFLAI ResearchRead Article

Quantitative Digital Microscopy with Deep Learning

IFLAI ResearchFebruary 10, 2021

Quantitative Digital Microscopy with Deep Learning

Our comprehensive 2021 Applied Physics Reviews paper outlining the transition from classical image processing to deep learning in digital microscopy.

IFLAI ResearchRead Article