Histogram-equalized quantization for logic-gated residual neural networks

Van Thien Nguyen (1) , William Guicquero (1, 2) , Gilles Sicard (1, 2)

1 DOPT - Département d'Optronique
2 DACLE-LETI - Département d'Architectures, Conception et Logiciels Embarqués-LETI

Adjusting the quantization according to the data or to the model loss seems mandatory to enable a high accuracy in the context of quantized neural networks. This work presents Histogram-Equalized Quantization (HEQ), an adaptive framework for linear and symmetric quantization. HEQ automatically adapts the quantization thresholds using a unique step size optimization. We empirically show that HEQ achieves state-of-the-art performances on CFAR-10. Experiments on the STL-10 dataset even show that HEQ enables a proper training of our proposed logic-gated (OR, MUX) residual networks with a higher accuracy at a lower hardware complexity than previous work.

Mots clés: CNN; quantized neural networks; histogram equalization; skip connections; logic-gated CNN

Domaines: Sciences de l'ingénieur [physics]

Van Thien Nguyen, William Guicquero, Gilles Sicard. Histogram-equalized quantization for logic-gated residual neural networks. ISCAS 2022 - 2022 IEEE International Symposium on Circuits and Systems, May 2022, Austin, France. pp.1289-1293, ⟨10.1109/ISCAS48785.2022.9937290⟩. ⟨cea-04556166⟩