Embedded Files
The GenΞΌ 2.0 Challenge @ U&ME 2026 Workshop
Challenge Description: As part of the workshop, this year also, we are organizing the Generative Machine Unlearning Challenge (GenΞΌ 2.0) to benchmark and advance research in generative machine unlearning and model editing. The 2026 update expands the original GenΞΌ concept-unlearning benchmark into a broader privacy-preserving suite co-located with the U&ME Workshop at ECCV 2026. Three tracks test whether a generative model can forget one specified identity or concept without damaging nearby identities, attributes, styles, or general generation quality. Check out this link to participate and find for more details on the challenge.
List of submissions that won the GenMU 2.0 challenge:
Β π Face Track:Β
Team name: SPRINT
Team: Fahad Shamshad, Mohammed Talha Alam, Fakhri Karray, Karthik Nandakumar
Paper: MARS: Manifold-Aware Anchor Redirection for Selective Face Identity Unlearning
π Speech Track:
Team name: FAA!
Team: Fahad Shamshad, Mohammed Talha Alam, Fakhri Karray, Karthik Nandakumar
Paper: Speaker-Embedding Gating for Voice-Cloning TTS: Speaker Identity Unlearning in XTTS-v2
π Β Visual Concepts Track:
Team name: Coders
Team: Ashok Urlana, L D M S Sai Teja, Vivek Hruday Kavuri, Ponnurangam Kumaraguru
Paper Title: MapRoute++: Surrogate-Guided Semantic Routing for Visual Concept Unlearning
This paper is accepted in the workshop also.
Page updated
Google Sites
Report abuse