Posts by Collection

ASL Letter Recognition

GUI using CNN and YOLO for American Sign Language verification

GitHub TensorFlow

Aqueous Solubility Prediction

Graph neural network for predicting aqueous solubility classification

PyTorch Kaggle

GPT4Readability: AI-Powered Code Documentation

AI-powered CLI tool for automatic code documentation

GitHub HuggingFace LangChain

Depth-Anything Video Analysis

Monocular depth estimation for autonomous UAVs

HuggingFace PyTorch KEF Robotics

Face Verification System

End-to-end face verification with CNNs

GitHub PyTorch

Digital Twins with Gaussian Splatting

Pipeline for creating Digital Twins of objects through Gaussian Splatting

PyTorch KEF Robotics

ML Models from Scratch

Machine learning models built exclusively with NumPy

GitHub NumPy

Jupyter Notebook Search Package

Python package for searching Jupyter notebooks

GitHub PyPI Python

Patient Stay Duration Prediction

DNN for predicting hospital stay duration

GitHub Presentation PyTorch

Question Answering Model

Question answering system using SpaCy

GitHub SpaCy

Rhine Water Level Prediction

VAR model for predicting Rhine river water levels

GitHub Time-Series

SkinsAI: Melanoma Detection Tool

Deep learning tool for melanoma detection (2nd place, Pitt Challenge)

Website Devpost PyTorch Django

Speech Phoneme Recognition

Phoneme state label identification for variable-length utterances

GitHub PyTorch

Speech-to-Text Transcription

RNN/BLSTM with dense networks for speech transcription

GitHub PyTorch

Stereo Disparity Map Generation

State-of-the-art disparity map generation using PSMNet

GitHub PyTorch

Vision-Language Model Object Counting

Enumerating strategies to improve VLM counting and investigating internal mechanisms

PyTorch Research

Improving LLM SVG Generation with Reinforcement Learning

Using GRPO to improve VLM SVG generation quality

Report PyTorch

Resolving Electrocatalytic Imprecision in Atomically Precise Metal Nanoclusters

Published in Current Opinion in Chemical Engineering, 2022

Review article on nanoparticle electrocatalyst design, covering machine learning applications and background in computational chemistry.

Recommended citation: Anantha Venkataraman Nagarajan, Dennis Johan Loevlie, Michael J Cowan, and Giannis Mpourmpakis. (2022). "Resolving electrocatalytic imprecision in atomically precise metal nanoclusters." Current Opinion in Chemical Engineering. 36:100784.
Download Paper

Size-dependent Shape Distributions of Platinum Nanoparticles

Published in Nanoscale Advances, 2022

Collaborative work applying Boltzmann statistics to explain experimental findings about platinum nanoparticle shape distributions.

Recommended citation: Ruikang Ding, Ingrid M. Padilla Espinosa, Dennis Loevlie, Soodabeh Azadehranjbar, Andrew J. Baker, Giannis Mpourmpakis, Ashlie Martini, and Tevis D. B. Jacobs. (2022). "Size-dependent shape distributions of platinum nanoparticles." Nanoscale Advances. 4:3978–3986.
Download Paper

Demystifying the Chemical Ordering of Multimetallic Nanoparticles

Published in Accounts of Chemical Research, 2023

A novel method to predict the chemical ordering of multimetallic nanoparticles, achieving a 71% reduction in RMSE compared to previous approaches.

Recommended citation: Dennis Johan Loevlie, Brenno Ferreira, and Giannis Mpourmpakis. (2023). "Demystifying the chemical ordering of multimetallic nanoparticles." Accounts of Chemical Research. 56(3):248–257.
Download Paper

Single Atom Alloys Segregation in the Presence of Ligands

Published in The Journal of Physical Chemistry C, 2023

Investigating how ligands affect single atom alloy segregation using machine learning and computational chemistry.

Recommended citation: Salem, M., Loevlie, D. J., Mpourmpakis, G. (2023). "Single Atom Alloys Segregation in the Presence of Ligands." The Journal of Physical Chemistry C. 127(46):22790-22798.
Download Paper

Synthetic Data Reveals Generalization Gaps in Correlated Multiple Instance Learning

Published in ML4H 2025 Symposium, Findings Track, 2025

We evaluate how well current MIL models capture inter-instance correlations by generating synthetic datasets and constructing a Bayes estimator as an optimal upper bound on model performance.

Recommended citation: Ethan Harvey, Dennis Johan Loevlie, and Michael C. Hughes. (2025). "Synthetic Data Reveals Generalization Gaps in Correlated Multiple Instance Learning." ML4H 2025 Symposium, Findings Track.
Download Paper

An Interactive Exploration of Computer Vision for Unmanned Aerial Vehicles (UAVs)

Teaching experience 1

Undergraduate course, University 1, Department, 2014

This is a description of a teaching experience. You can use markdown like any other post.

Teaching experience 2

Workshop, University 1, Department, 2015

This is a description of a teaching experience. You can use markdown like any other post.