Bit-ViP applies bit-plane processing with Lorenz chaotic noise and differential privacy to produce obfuscated images that resist reconstruction attacks yet support activity recognition on UCF101 and HMDB51.
Differentially private image classification from features,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Bit-ViP: Leveraging Bit-planes to Preserve Visual Privacy in Images through Obfuscation
Bit-ViP applies bit-plane processing with Lorenz chaotic noise and differential privacy to produce obfuscated images that resist reconstruction attacks yet support activity recognition on UCF101 and HMDB51.