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

Learning-based pre-simulation (i.e., layout-to-fabrication) models have beenproposed to predict the fabrication-induced shape deformation from an IC layoutto its fabricated circuit. Such models are usually driven by pairwise learning,involving a training set of layout patterns and their reference shape imagesafter fabrication. However, it is expensive and time-consuming to collect thereference shape images of all layout clips for model training and updating. Toaddress the problem, we propose a deep learning-based layout novelty detectionscheme to identify novel (unseen) layout patterns, which cannot be wellpredicted by a pre-trained pre-simulation model. We devise a global-localnovelty scoring mechanism to assess the potential novelty of a layout byexploiting two subnetworks: an autoencoder and a pretrained pre-simulationmodel. The former characterizes the global structural dissimilarity between agiven layout and training samples, whereas the latter extracts a latent coderepresenting the fabrication-induced local deformation. By integrating theglobal dissimilarity with the local deformation boosted by a self-attentionmechanism, our model can accurately detect novelties without the ground-truthcircuit shapes of test samples. Based on the detected novelties, we furtherpropose two active-learning strategies to sample a reduced amount ofrepresentative layouts most worthy to be fabricated for acquiring theirground-truth circuit shapes. Experimental results demonstrate i) our method'seffectiveness in layout novelty detection, and ii) our active-learningstrategies' ability in selecting representative novel layouts for keeping alearning-based pre-simulation model updated.

Keeping Deep Lithography Simulators Updated: Global-Local Shape-Based Novelty Detection and Active Learning