Poster
Latent Weights Do Not Exist: Rethinking Binarized Neural Network Optimization
Koen Helwegen · James Widdicombe · Lukas Geiger · Zechun Liu · Kwang-Ting Cheng · Roeland Nusselder

Wed Dec 11th 10:45 AM -- 12:45 PM @ East Exhibition Hall B + C #103

Optimization of Binarized Neural Networks (BNNs) currently relies on real-valued latent weights to accumulate small update steps. In this paper, we argue that these latent weights cannot be treated analogously to weights in real-valued networks. Instead their main role is to provide inertia during training. We interpret current methods in terms of inertia and provide novel insights into the optimization of BNNs. We subsequently introduce the first optimizer specifically designed for BNNs, Binary Optimizer (Bop), and demonstrate its performance on CIFAR-10 and ImageNet. Together, the redefinition of latent weights as inertia and the introduction of Bop enable a better understanding of BNN optimization and open up the way for further improvements in training methodologies for BNNs.

Author Information

Koen Helwegen (Plumerai)
James Widdicombe (Plumerai)
Lukas Geiger (Plumerai)
Zechun Liu (HKUST)
Kwang-Ting Cheng (Hong Kong University of Science and Technology)
Roeland Nusselder (Plumerai)