Get Started
Research questionHow can counterfactual image edits be audited for a shared latent explanation across regions?An edited image can satisfy plausibility checks for each region while the regional outputs are mutually incompatible with any single latent explanation. Auditing must therefore distinguish local plausibility from global coherence and determine what can be inferred about declared image features.
Computer Vision
Image & Video Processing
Machine Learning
Latest papersRecent research connected to this question, newest first.Common-Witness Certificates and Sharp Feature Bounds for Counterfactual Image AuditingApplies to the common-witness framework for image editors, evaluated on controlled MNIST, Morpho-MNIST, and smallNORB studies. The evidence concerns externally justified witness relations, sharp bounds for declared features, finite-sample outer coverage, and exact repair counts; it does not establish unrestricted pixel-level counterfactuals.research paper · Sep 3, 2026
Related questions
How can image editors infer edit regions and preserve non-target content without introducing artifacts?How can we generate realistic, minimally edited visual counterfactuals without inheriting classifier shortcuts?How can medical-image counterfactuals reveal a classifier’s decisions without inheriting a generator’s biases?How can text-guided image editing explain edits while keeping provenance resistant to white-box removal?
Home
Topics
Search
Library