Quantization of Deep Neural Networks for Accumulator-constrained Processors

no code implementations • 24 Apr 2020 • Barry de Bruin, Zoran Zivkovic, Henk Corporaal

We demonstrate that 16-bit accumulators are able to obtain a classification accuracy within 1\% of the floating-point baselines on the CIFAR-10 and ILSVRC2012 image classification benchmarks.

Classification General Classification +2