Maps frozen MobileNetV2 features to Ising spins on quasi-cyclic LDPC graphs, operates a Random-Bond Ising Model at Nishimori temperature, and achieves 98.7% top-1 accuracy on ImageNet-10 and 84.92% on ImageNet-100 with 32-64 dimensional representations.
In particular, modeling feature vectors as spins (nodes) on a sparse graph under a Gibbs distribu tion (a Random-Bond Ising Model) enables powerful clustering and classification
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Natural Image Classification via Quasi-Cyclic Graph Ensembles and Random-Bond Ising Models at the Nishimori Temperature
Maps frozen MobileNetV2 features to Ising spins on quasi-cyclic LDPC graphs, operates a Random-Bond Ising Model at Nishimori temperature, and achieves 98.7% top-1 accuracy on ImageNet-10 and 84.92% on ImageNet-100 with 32-64 dimensional representations.