Wisdom O. Ikezogwo

Ph.D. Candidate in Computer Science & Engineering | AI/ML Research Scientist

Seattle, US.

About

Highly accomplished Ph.D. Candidate in Computer Science & Engineering with a strong track record of leading innovative research in Generative AI, Multimodal Representation Learning, and Data Curation. Expert in developing state-of-the-art machine learning models for complex domains including medical imaging, egocentric data, and financial analytics. Proven ability to drive projects from conception to clinical/production evaluation, mentor teams, and publish extensively in top-tier conferences and journals. Seeking to leverage advanced AI/ML expertise to solve challenging real-world problems and contribute to cutting-edge technological advancements.

Work

University of Washington

Research Assistant, Graphics and Imaging Laboratory (GRAIL)

Summary

Leading advanced research in Multimodal Large Language Models (LLMs) for medical imaging and developing cutting-edge generative models.

Highlights

Spearheaded research in Multimodal Large Language Models (LLMs) for medical imaging, developing novel medical multimodal datasets (Quilt-1M, MedNarratives) and state-of-the-art analytical models (QuiltNet, Quilt-LLaVA).

Created a multi-agent AI framework (PathFinder) for clinical diagnosis that outperformed human experts, and established improved benchmarks (MedBlink).

Directing initiatives to enhance physics-informed image and video generative models, focusing on Newtonian physics for the generation of large-scale video scene graph datasets and pipelines.

Apple

Ph.D. Machine Learning Research Intern

Summary

Led research on efficient multimodal representations for egocentric data, optimizing data capture costs.

Highlights

Spearheaded research into efficient multimodal representations for egocentric data (video, text, audio, IMU, hands), developing the "Perceive-Predict" model.

Utilized predictive coding between co-occurring modalities to reconstruct missing modalities, significantly reducing data capture costs for expensive modalities like video.

Mayo Clinic

Ph.D. Quantitative Health Sciences Intern

Summary

Directed research for a foundational AI model in histopathology, scaling computational efforts and validating models clinically.

Highlights

Directed research and development of a foundational AI model for histopathology, leveraging millions of gigapixel histology images.

Scaled computational efforts on the Argonne National Lab computing cluster and conducted rigorous clinical evaluations to validate model efficacy.

Okra, Inc.

ML Engineer

Summary

Developed and deployed machine learning models for financial data analysis to enhance lending processes.

Highlights

Engineered and deployed machine learning models to analyze unstructured banking data, extracting critical customer earning and spending insights.

Instrumental in enhancing downstream lending processes, including income prediction, spending pattern analysis, and financial reconciliations.

Demz Analytics Limited

Data Scientist / ML Engineer

Summary

Designed and implemented production-grade recommendation systems to optimize user engagement.

Highlights

Designed and implemented production-grade recommendation systems, leveraging advanced techniques such as attention mechanisms and epsilon-greedy bandit strategies.

Optimized system performance and user engagement through data-driven model development and deployment.

Obafemi Awolowo University

UG. Research Assistant, Biosignal Processing, Inst. & Control Lab

Summary

Conducted research on biosignal processing and neural networks for brain EEG signal analysis.

Highlights

Integrated and analyzed disparate multivariate time series data, focusing on characterizing spectral components for dynamic dimensionality reduction.

Developed and trained neural networks for the classification of brain EEG signals and comprehensive characterization of EEG spectral components.

Education

University of Washington

Ph.D. Computer Science and Engineering Grade: 3.97/4.00

Courses

Generative Modeling

Multimodal Representation Learning

Deep Learning

Computer Vision

Natural Language Processing

Artificial Intelligence

Obafemi Awolowo University

B.Sc. Electronic & Electrical Engineering Grade: 4.73/5.00 (Class rank 2/120)

Courses

Digital Signal Processing

Neural Networks

Control Systems

Embedded Systems

Electromagnetics

Awards

Population Health Initiative - AI Pilot Research Grant Award

University of Washington

Awarded $100,000 for pioneering AI research in population health.

Microsoft's Accelerate Foundation Models Research Grant

Microsoft

Received $20,000 grant to advance research in foundational AI models.

IBRO-Simons Computational Neuroscience Summer School Travel Grant

IBRO-Simons

Awarded travel grant to attend the Computational Neuroscience Summer School in Cape Town.

Prof. Kehinde Prize for the Best Graduating Student in Control Option

Obafemi Awolowo University

Recognized as the top-performing student in the Control Engineering specialization.

Oyebolu Prize for Best Male Graduating Student

Obafemi Awolowo University

Awarded for outstanding academic achievement as the best male graduating student.

Federal Government Scholarship Award, Nigeria

Federal Government of Nigeria

National scholarship awarded for academic excellence (cumulative value: $1500).

Total/NNPC National Merit Scholarship

Total/NNPC

Merit-based scholarship awarded for academic achievement (cumulative value: $1500).

Etisalat Nigeria Merit Scholarship

Etisalat Nigeria

Merit-based scholarship awarded for academic achievement (value: $250).

Publications

Quilt-LLaVA: Visual Instruction Tuning by Extracting Localized Narratives from Open-Source Histopathology Videos.

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

Contributed to research on visual instruction tuning for histopathology videos using localized narratives.

Quilt-1M: One Million Image-Text Pairs for Histopathology.

NeurIPS

Led the development of a large-scale image-text dataset for histopathology, presented as an oral paper.

Multi-modal Masked Autoencoders Learn Compositional Histopathological Representations.

Machine Learning for Health (ML4H)

Authored an extended abstract on learning compositional representations for histopathology using masked autoencoders.

Risk Stratification of Solitary Fibrous Tumor Using Whole Slide Image Analysis.

LABORATORY INVESTIGATION, ELSEVIER SCIENCE INC

Contributed to research on risk stratification for solitary fibrous tumors using whole slide image analysis.

Recent Advances, Applications and Open Challenges in Machine Learning for Health: Reflections from Research Roundtables

ML4H Symposium

Contributed to a collaborative paper reflecting on recent advances and challenges in Machine Learning for Health.

Supervised domain generalization for integration of disparate scalp EEG datasets for automatic epileptic seizure detection.

Computers in Biology and Medicine

Co-authored research on domain generalization for EEG seizure detection across disparate datasets.

Empirical Characterization of the Temporal Dynamics of EEG Spectral Components.

International Journal of Online and Biomedical Engineering (IJOE)

Co-authored research on empirical characterization of temporal dynamics in EEG spectral components.

Synthetic Video Scene Graph Generation

NeurIPS D&B

Forthcoming publication on generating synthetic video scene graphs.

Multi-Scale Cross-Attention Multiple Instance Learning (MsCAMIL) Network for Automated Triage of Colorectal Polyps.

United States and Canadian Academy of Pathology's (USCAP) 114th Annual Meeting

Forthcoming presentation on a novel network for automated triage of colorectal polyps.

Comparative Performance of Multi-Scale Cross-Attention Multiple Instance Learning (MsCAMIL) and Pathology Trainees in Colorectal Polyp Diagnosis.

United States and Canadian Academy of Pathology's (USCAP) 114th Annual Meeting

Forthcoming comparative study on MsCAMIL performance against pathology trainees in polyp diagnosis.

PathFinder: A Multi-Modal Multi-Agent Framework for Diagnostic Decision-Making in Histopathology.

ICCV

In submission: A multi-modal, multi-agent framework for diagnostic decision-making in histopathology.

MedicalNarratives: Connecting Medical Vision and Language with Procedural and Localized Narratives across all medical imaging domains.

ICCV

In submission: Research on connecting medical vision and language through narratives across imaging domains.

MedBlink: Probing the Fundamental Medical Imaging Knowledge of Multimodal Language Models.

ICCV

In submission: Research exploring the fundamental medical imaging knowledge within multimodal language models.

Percieve-Predict: Modality and Time-Aware Egocentric Efficient Multi-Modal Representations.

NeurIPS

In preparation: Research on efficient multi-modal representations for egocentric data.

VPhysics: Temporally consistent Physics in Video (multiframe) Generation via Alignment

NeurIPS

In preparation: Research on generating temporally consistent physics in video through alignment.

Languages

English

Native

Skills

Machine Learning & AI

Generative Modeling, Multimodal Representation Learning, Large Language Models (LLMs), Deep Learning, Neural Networks, Computer Vision, Natural Language Processing (NLP), Reinforcement Learning, Predictive Modeling, Histopathology AI, Biosignal Processing, Time Series Analysis, Recommendation Systems, Attention Mechanisms, Bandit Algorithms.

Programming & Tools

Python, PyTorch, TensorFlow, Data Curation, Data Analysis, SQL, Cloud Computing (Argonne National Lab Cluster), Git.

Research & Development

Scientific Writing, Experimental Design, Model Evaluation, Statistical Analysis, Data Collection, Peer Review, Algorithm Development, Prototyping.

Teaching & Leadership

Teaching Assistant, Mentorship, Curriculum Development, Project Leadership, Team Collaboration, Problem Solving, Critical Thinking, Communication.

Projects

PathFinder Multi-Agent AI Framework

Summary

Designed and implemented a multi-agent AI framework for clinical diagnosis in histopathology.

Quilt-1M & MedNarratives Datasets

Summary

Developed large-scale medical multimodal datasets (Quilt-1M: 1M Image-Text Pairs, and MedNarratives) to advance research in AI for histopathology.

Perceive-Predict Model

Summary

Developed an efficient multimodal representation model for egocentric data (video, text, audio, IMU, hands).

Foundational Model for Histopathology

Summary

Led research and development of a foundational AI model for histopathology.

EEG Seizure Detection Application

Summary

Developed an application for combining disparate EEG seizure datasets into a single, unified dataset.