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

Human pose transfer synthesizes new view(s) of a person for a given pose.Recent work achieves this via self-reconstruction, which disentangles aperson's pose and texture information by breaking the person down into parts,then recombines them for reconstruction. However, part-level disentanglementpreserves some pose information that can create unwanted artifacts. In thispaper, we propose Pose Transfer by Permuting Textures (PT$^2$), an approach forself-driven human pose transfer that disentangles pose from texture at thepatch-level. Specifically, we remove pose from an input image by permutingimage patches so only texture information remains. Then we reconstruct theinput image by sampling from the permuted textures for patch-leveldisentanglement. To reduce noise and recover clothing shape information fromthe permuted patches, we employ encoders with multiple kernel sizes in a triplebranch network. On DeepFashion and Market-1501, PT$^2$ reports significantgains on automatic metrics over other self-driven methods, and even outperformssome fully-supervised methods. A user study also reports images generated byour method are preferred in 68% of cases over self-driven approaches from priorwork. Code is available at https://github.com/NannanLi999/pt_square.

Collecting The Puzzle Pieces: Disentangled Self-Driven Human Pose Transfer by Permuting Textures