Deep Leakage from Gradients

Zhu, L., Liu, Z., & Han, S. Deep Leakage from Gradients.

Type

Book chapter

Intellectual base

Review or monograph

Link

https://arxiv.org/abs/1906.08935

Cited from

computer_science

Cited by

1 citation across 1 artifact.

Each citation links to the sentence it supports in the citing article.

Primes

• Hidden Information Reconstruction
  • With a sharp image prior, shared gradients reconstruct contributing images at near-pixel fidelity — defeating the "data never leaves the device" framing exactly as the prime predicts.

    This sourceIn Advances in Neural Information Processing Systems (NeurIPS) 32. 2019. Reconstructs a client's training images at near-pixel fidelity from shared gradients under a natural-image prior, defeating the 'data never leaves the device' framing of federated learning — gradient inversion.

Verification

This reference passed the adversarial substantiation pipeline: it was checked to exist and to support the claim it is attached to. See how references were verified.

Registry ID ref:fae8b31038d8 · see in the full table