BEGIN:VCALENDAR VERSION:2.0 PRODID:Linklings LLC BEGIN:VTIMEZONE TZID:Asia/Seoul X-LIC-LOCATION:Asia/Seoul BEGIN:STANDARD TZOFFSETFROM:+0900 TZOFFSETTO:+0900 TZNAME:KST DTSTART:18871231T000000 DTSTART:19881009T020000 END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20230103T035307Z LOCATION:Auditorium\, Level 5\, West Wing DTSTART;TZID=Asia/Seoul:20221206T100000 DTEND;TZID=Asia/Seoul:20221206T120000 UID:siggraphasia_SIGGRAPH Asia 2022_sess153_papers_436@linklings.com SUMMARY:Reference Based Sketch Extraction via Attention Mechanism DESCRIPTION:Technical Papers\n\nReference Based Sketch Extraction via Atte ntion Mechanism\n\nAshtari, Seo, Kang, Cha, Noh\n\nWe propose a model that extracts a sketch from a colorized image in such a way that the extracted sketch has a line style similar to a given reference sketch while preserv ing the visual content identically to the colorized image. Authentic sketc hes drawn by artists have various sketch styles to add visual interest and contribute feeling to the sketch. However, existing sketch-extraction met hods generate sketches with only one style. Moreover, existing style trans fer models fail to transfer sketch styles because they are mostly designed to transfer textures of a source style image instead of transferring the sparse line styles from a reference sketch. Lacking the necessary volumes of data for standard training of translation systems, at the core of our G AN-based solution is a self-reference sketch style generator that produces various reference sketches with a similar style but different spatial lay outs. We use independent attention modules to detect the edges of a colori zed image and reference sketch as well as the visual correspondences betwe en them. We apply several loss terms to imitate the style and enforce spar sity in the extracted sketches. Our sketch-extraction method results in a close imitation of a reference sketch style drawn by an artist and outperf orms all baseline methods. Using our method, we produce a synthetic datase t representing various sketch styles and improve the performance of auto-c olorization models, in high demand in comics. The validity of our approach is confirmed via qualitative and quantitative evaluations.\n\nRegistratio n Category: FULL ACCESS, EXPERIENCE PLUS ACCESS, EXPERIENCE ACCESS, TRADE EXHIBITOR\n\nLanguage: ENGLISH\n\nFormat: IN-PERSON URL:https://sa2022.siggraph.org/en/full-program/?id=papers_436&sess=sess15 3 END:VEVENT END:VCALENDAR