Gallery
CONFERENCES / MILESTONES / STARTUP / RESEARCH
A source-linked visual record of research and practice. Conference material is grouped by event so posters, presentation photographs and related visuals stay together.
Public profiles
MILESTONE
Profile & milestones
Public profile photographs and academic milestones.
Profile portrait.
Safe & Trusted AI feature marking PhD completion and the next stage of my research.
CONFERENCE
NeurIPS 2024
Identifiable Object-Centric Representation Learning via Probabilistic Slot Attention.
Visual shared with my NeurIPS 2024 acceptance announcement.
Second visual from the NeurIPS 2024 announcement.
Research visual for probabilistic slot attention and identifiable object-centric representations.
CONFERENCE
ICML 2025 · Vancouver
Three papers spanning object-centric representation learning, causal discovery and diffusion counterfactuals.
Identifiable Object Representations under Spatial Ambiguities — multi-view object-centric learning under occlusion and viewpoint ambiguity.
Continuous Bayesian Model Selection for Multivariate Causal Discovery — flexible causal discovery beyond restrictive functional assumptions.
Diffusion Counterfactual Generation with Semantic Abduction — causal control with identity-preserving diffusion mechanisms.
CONFERENCE
MICCAI 2019 · Shenzhen
Presentation, poster sessions and community moments from earlier medical-imaging work.
Presenting research at MICCAI 2019.
Poster session and research discussions.
Research networking during MICCAI 2019.
MICCAI 2019 conference gala.
MICCAI 2019, Shenzhen.
Shenzhen during the conference trip.
STARTUP
Blurgs AI · maritime & startup milestones
Publicly documented company milestones and maritime-intelligence work.
Blurgs AI founders with the maritime-intelligence platform.
Public founder photograph accompanying the 2026 funding milestone.
Startup Maharathi 2025: Blurgs' Defence & Space Tech recognition at Bharat Mandapam, New Delhi.
A second photograph from Blurgs' Startup Maharathi 2025 recognition in New Delhi.
RESEARCH
Selected research visuals
Visuals from earlier explainability, medical-imaging and computational-neuroscience projects.
BioExp: an early explainability pipeline for biomedical deep-learning models.
Model-attribution visual from the BioExp project.
Brain-tumour segmentation overlay from earlier medical-imaging work.
DigiPathAI: deep-learning tools for computational pathology.
Neural-field modelling from earlier computational-neuroscience work.
No gallery groups match this filter.
Only identity-matched public profiles and clearly attributable public images are included. Search results for other people with the same name are intentionally excluded.