Geminio uses a vision-language model to reshape a federated model so that gradients from a victim client reveal only the images that match an attacker's text query.
Challenges in representation learning: Fa- cial expression recognition challenge
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Geminio: Language-Guided Gradient Inversion Attacks in Federated Learning
Geminio uses a vision-language model to reshape a federated model so that gradients from a victim client reveal only the images that match an attacker's text query.