DASM is a new optimizer integrating domain-supervised contrastive learning and sharpness-aware minimization with adaptive gap modulation to boost generalization and robustness in multi-domain voice stream steganalysis.
Frame-level steganalysis of qim steganog- raphy in compressed speech based on multi-dimensional perspective of codeword correlations,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
DASM: Domain-Aware Sharpness Minimization for Multi-Domain Voice Stream Steganalysis
DASM is a new optimizer integrating domain-supervised contrastive learning and sharpness-aware minimization with adaptive gap modulation to boost generalization and robustness in multi-domain voice stream steganalysis.