Computer Vision · ML · Explainable AI

I build visual intelligence that explains itself.

Researching interpretable machine learning for complex scenes, where computer vision, human understanding, and interactive AI meet.

Based atOregon State University
Research focusInterpretable ML
Open toResearch collaboration
Bhavan Vasu presenting research at an academic event
INTERPRETABILITY
HUMAN × AI
VISUAL REASONING
RELATIONAL LEARNING
Scroll to discover
01

Selected work

Research made tangible.

A selection of projects spanning video explanations, relational learning, global and local explanations, natural-language interfaces, content-based retrieval, planetary AI, medical imaging, aerial imagery, and visual reasoning.

GRADIENT → LLM → RULES LLM 0.74 : toxic(X) ← bond(X,Y) + 0.31 : cites(A,B) ← ... RELATIONAL LEARNING · IJCLR 2026
Relational learning · Neurosymbolic AI

GRASP: LLM-Guided Functional Gradient Boosting

A neurosymbolic approach that uses functional gradients to guide an LLM toward useful first-order rules, replacing combinatorial symbolic search while retaining interpretable probabilistic logic programs.

A ∧ B ∨ C LOCAL → GLOBAL
Logical explanations

Local-to-Global Logical Explanations

Human-recognizable primitive concepts are composed into logical formulas that explain individual predictions, classes, and multiclass model behavior.

IJCLR 2025 ↗
CONCEPTS ACROSS A CLASS
Global concept XAI

Beyond Local Explanations

A model-agnostic framework for explaining image classifiers globally with human-aligned concepts rather than only per-image attribution maps.

Paper ↗
Planetary AI · Retrieval

Interactive Mars Image Search

Prototype-based, interpretable content search for NASA's Planetary Data System, designed to help users inspect and validate evidence behind Mars image classifications.

Paper ↗
SIMILARITY + SALIENCY
Medical XAI · Retrieval

X-MIR: Explainable Medical Image Retrieval

Explainable retrieval for chest X-rays and skin-lesion imagery, using saliency to reveal which visual evidence drives similarity between medical images.

Paper ↗
{ road, car } SUFFICIENT CONCEPT SET
Scene understanding · XAI

Global Explanations for Image Classifiers

Explaining complex-scene classifiers with compact sets of relevant objects, using search and dimensionality reduction to summarize model behavior locally and globally.

AAAI ↗
HUMAN ↔ MODEL
Human-in-the-loop retrieval

Explainable, Interactive Image Retrieval

A utility-based approach to measuring whether explanations actually help people perform a retrieval task, with similarity saliency integrated into an interactive search loop.

Paper ↗
Diagram comparing conventional machine learning and explainable AI
DARPA XAI

Explainable Artificial Intelligence

Research toward systems that communicate why they made a decision—not only what they predicted.

Program ↗
Visualization from research on neural network interpretation resilience
Aerial imagery

Resilience of Deep Network Interpretations

Investigating how model interpretations behave under perturbations in satellite and aerial imagery.

Thesis ↗
Dual-head neural network architecture for reading jersey numbers
Joint learning

Jersey Number Reader

A dual-head joint-learning architecture for reading two-digit jersey numbers from imagery.

Code ↗
Siamese network architecture for image change detection
Remote sensing

Bi-temporal Change Detection

Siamese-style neural architectures for finding meaningful change across paired satellite images.

Code ↗
Illustrative artwork about probabilistically checkable proofs
Theoretical CS

Probabilistically Checkable Proofs

An accessible exploration of PCPs, their surprising power, and the weak PCP theorem.

Diagram illustrating metric learning neighborhoods
Learning theory

Generalization Bounds for Metric Learning

Exploring Rademacher complexity and matrix norms to characterize generalization in metric and similarity learning.

02

Publications

Ideas, peer reviewed.

Peer-reviewed papers and conference work, alongside the latest Scholar-listed preprints in interpretable vision, video understanding, relational learning, retrieval, explainability, and remote sensing.

Google Scholar ↗

Disentangling Where and When: Factored Spatio-Temporal Explanations for Video Action Recognition

NeurIPS 2026 · Bhavan Vasu, Giuseppe Raffa, Prasad Tadepalli

Open ↗

Learning Probabilistic Logic Programs with Functional Gradient Guided Language Models

IJCLR 2026 · Saurabh Mathur, Sahil Sidheekh, Bhavan Vasu, Farbod Tavakkoli, Prasad Tadepalli, Kristian Kersting, Sriraam Natarajan

Open ↗

GLARE: A Natural Language Interface for Querying Global Explanations

Preprint · Bhavan Vasu, Rajesh Mangannavar

Open ↗

Local-to-Global Logical Explanations for Deep Vision Models

IJCLR 2025 · Bhavan Vasu, Giuseppe Raffa, Prasad Tadepalli

Open ↗

Beyond Local Explanations: A Framework for Global Concept-Based Interpretation in Image Classification

Electronics · Bhavan Vasu, Kunal Rathore, Prasad Tadepalli

Open ↗

Interactive Mars Image Content-Based Search with Interpretable Machine Learning

IAAI 2024 · Bhavan Vasu, Steven Lu, Emily Dunkel, Kiri L. Wagstaff, Kevin Grimes, Michael McAuley

Open ↗

Global Explanations for Image Classifiers

AAAI Student Abstract Program · Bhavan Vasu, Prasad Tadepalli

AAAI

Content-based Search of Large Image Archives at PDS Imaging Node

American Geophysical Union · Steven Lu, Kiri Wagstaff, Emily Dunkel, Bhavan Vasu, Kevin Grimes, Michael McAuley

AGU

X-MIR: EXplainable Medical Image Retrieval

WACV · Brian Hu, Bhavan Vasu, Anthony Hoogs

Open ↗

XAITK: The Explainable AI Toolkit

Applied AI Letters · Brian Hu, Paul Tunison, Bhavan Vasu, Nitesh Menon, Roddy Collins, Anthony Hoogs

Open ↗

Explainable, Interactive Content-Based Image Retrieval

Applied AI Letters · Bhavan Vasu, Brian Hu, Bo Dong, Roddy Collins, Anthony Hoogs

Open ↗

Resilience and Plasticity of Deep Network Interpretations for Aerial Imagery

IEEE Access · Bhavan Vasu, Andreas Savakis

IEEE

Iterative and Adaptive Sampling with Spatial Attention for Black-Box Model Explanations

WACV · Bhavan Vasu, Chengjiang Long

Open ↗
03

Journey

Built across disciplines.

From electronics and imaging science to applied computer vision, explainability, and planetary data.

2011–2015

Education

B.E. in Electronics & Communication Engineering

Global Academy of Technology · Bangalore, India
2016–2018

Education + Research

M.S. in Computer Engineering

Rochester Institute of Technology · Center for Imaging Science
2018–2021

Industry research

R&D Engineer, Computer Vision

Kitware · New York
2022

Planetary AI

Graduate Fellowship Intern

NASA Jet Propulsion Laboratory · California
2021–Present

Doctoral research

Ph.D. in Computer Science

Oregon State University · advised by Prof. Prasad Tadepalli

About

Rigorous research.
Creative curiosity.

My work uses computer vision and machine learning to find patterns, inspect model behavior, and turn opaque predictions into understandable interactions. I enjoy entering unfamiliar research spaces and building on existing ideas with practical, creative improvements.

Beyond research, I care about access to education and minority empowerment. Photography and music shape how I observe, frame, and communicate technical ideas.

Interpretable MLHuman–AI InteractionInteractive MLComputer VisionImage RetrievalRemote Sensing

“Be bold enough to follow the obvious, but have the tenacity to go after the behemoth.”

Personal maxim
04

Field notes

Milestones & moments.

A selected stream of talks, awards, papers, and community work.

Second place at Global Futures Forum: AI Week

Recognized at Oregon State University for work on global explanations.

Oral presentation at IAAI 2024

Presented interactive Mars image search research in Vancouver.

IAAI paper accepted

Interactive Mars Image Content-Based Search with Interpretable Machine Learning.

AAAI Workflow Chair

Appointed to support the conference’s research community and operations.

“Most Impactful” poster

Rule-Based Explanations for Deep Networks recognized at the PPI Center meeting.

NASA JPL Fellowship

Selected for the Graduate Fellowship Program supporting the Mars 2020 mission.

Collaborator perspective

Trusted to go beyond the obvious.

“He combines a deep knowledge of machine learning techniques and research with an unparalleled willingness to branch out and try new things.”

William HicksEngineer, RAPIDS AI, NVIDIA

“Bhavan knows the foundations clearly enough to understand subtle mistakes and is capable enough to appreciate subtle but breakthrough ideas.”

Nilesh PandeyComputer Vision Engineer, Dolby

“It’s rare to come across individuals who excel in different domains, and Bhavan is indeed a rare find.”

Raviteja GundaData Scientist, Roku

Contact

Let’s make opaque systems a little more understandable.

For research collaborations, speaking, reviewing, or thoughtful conversations about visual intelligence.

Start a conversation

What would you like to explore?

Share a little context and your message will be delivered by email.