LAB: Learnable Activation Binarizer for Binary Neural Networks
This work addresses a bottleneck in BNNs for edge devices, offering an incremental improvement by enhancing information propagation with a novel module.
The paper tackles the limitation of using sign() for binarization in Binary Neural Networks by proposing a learnable activation binarizer (LAB) that replaces sign() with per-layer learnable kernels, resulting in a considerable performance boost when integrated into existing BNNs and achieving competitive state-of-the-art performance on ImageNet.
Binary Neural Networks (BNNs) are receiving an upsurge of attention for bringing power-hungry deep learning towards edge devices. The traditional wisdom in this space is to employ sign() for binarizing featuremaps. We argue and illustrate that sign() is a uniqueness bottleneck, limiting information propagation throughout the network. To alleviate this, we propose to dispense sign(), replacing it with a learnable activation binarizer (LAB), allowing the network to learn a fine-grained binarization kernel per layer - as opposed to global thresholding. LAB is a novel universal module that can seamlessly be integrated into existing architectures. To confirm this, we plug it into four seminal BNNs and show a considerable performance boost at the cost of tolerable increase in delay and complexity. Finally, we build an end-to-end BNN (coined as LAB-BNN) around LAB, and demonstrate that it achieves competitive performance on par with the state-of-the-art on ImageNet.