Updating the VESICLE-CNN Synapse Detector

TL;DR

This paper updates the VESICLE-CNN synapse detector to a fully convolutional, dilated convolution version, achieving the same accuracy as the original but with a 600-fold speed increase, and releases the code publicly.

Contribution

The paper introduces a fully convolutional, dilated convolution implementation of VESICLE-CNN, significantly improving speed without sacrificing accuracy.

Findings

600x faster test-time performance

Maintains original detection accuracy

Open-source release of code and data

Abstract

We present an updated version of the VESICLE-CNN algorithm presented by Roncal et al. (2014). The original implementation makes use of a patch-based approach. This methodology is known to be slow due to repeated computations. We update this implementation to be fully convolutional through the use of dilated convolutions, recovering the expanded field of view achieved through the use of strided maxpools, but without a degradation of spatial resolution. This updated implementation performs as well as the original implementation, but with a $600\times$ speedup at test time. We release source code and data into the public domain.