English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

For the past decade, convolutional networks have been used for 3Dreconstruction of neurons from electron microscopic (EM) brain images. Recentyears have seen great improvements in accuracy, as evidenced by submissions tothe SNEMI3D benchmark challenge. Here we report the first submission to surpassthe estimate of human accuracy provided by the SNEMI3D leaderboard. A variantof 3D U-Net is trained on a primary task of predicting affinities betweennearest neighbor voxels, and an auxiliary task of predicting long-rangeaffinities. The training data is augmented by simulated image defects. Thenearest neighbor affinities are used to create an oversegmentation, and thensupervoxels are greedily agglomerated based on mean affinity. The resultingSNEMI3D score exceeds the estimate of human accuracy by a large margin. Whileone should be cautious about extrapolating from the SNEMI3D benchmark toreal-world accuracy of large-scale neural circuit reconstruction, our resultinspires optimism that the goal of full automation may be realizable in thefuture.

Superhuman Accuracy on the SNEMI3D Connectomics Challenge