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 Plus annual

Presents the access of premium content as premium feature

Premium Content

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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

We introduce MeshGPT, a new approach for generating triangle meshes thatreflects the compactness typical of artist-created meshes, in contrast to densetriangle meshes extracted by iso-surfacing methods from neural fields. Inspiredby recent advances in powerful large language models, we adopt a sequence-basedapproach to autoregressively generate triangle meshes as sequences oftriangles. We first learn a vocabulary of latent quantized embeddings, usinggraph convolutions, which inform these embeddings of the local mesh geometryand topology. These embeddings are sequenced and decoded into triangles by adecoder, ensuring that they can effectively reconstruct the mesh. A transformeris then trained on this learned vocabulary to predict the index of the nextembedding given previous embeddings. Once trained, our model can beautoregressively sampled to generate new triangle meshes, directly generatingcompact meshes with sharp edges, more closely imitating the efficienttriangulation patterns of human-crafted meshes. MeshGPT demonstrates a notableimprovement over state of the art mesh generation methods, with a 9% increasein shape coverage and a 30-point enhancement in FID scores across variouscategories.

MeshGPT: Generating Triangle Meshes with Decoder-Only Transformers