BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Asia/Hong_Kong
X-LIC-LOCATION:Asia/Hong_Kong
BEGIN:STANDARD
TZOFFSETFROM:+0800
TZOFFSETTO:+0800
TZNAME:HKT
DTSTART:19911015T033000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20251218T030657Z
LOCATION:Meeting Room S423+S424\, Level 4
DTSTART;TZID=Asia/Hong_Kong:20251217T154400
DTEND;TZID=Asia/Hong_Kong:20251217T155500
UID:siggraphasia_SIGGRAPH Asia 2025_sess141_papers_2503@linklings.com
SUMMARY:FairyGen: Storied Cartoon Video from a Single Child-Drawn Characte
 r
DESCRIPTION:Jiayi Zheng (Great Bay University, Donghua University) and Xia
 odong Cun (Great Bay University)\n\nWe present FairyGen, an automatic syst
 em to generate storied cartoon videos from a single child's drawing charac
 ter with a highly personalized style. Unlike previous subjects and motion-
 customization methods, we identify the whole story as layers of character 
 modeling, environment generation, and shot design for the continuous story
 . Giving a single hand-drawn image,our approach initiates by utilizing the
  Multi-modality Large Language Model~(MLLM) to create a structured storybo
 ard that includes dynamic shots, setting up both the narrative flow and th
 e spatial layout of the main character. To model the character, we develop
  a 3D proxy that allows us to produce tailored motion that incorporates in
 tricate and real-world dynamics. Then, for environment generation, we desi
 gn a style propagation adapter to learn the style from the foreground char
 acter and propagate it to the background via the pre-trained background in
 painting diffusion models, so that the identity of the foreground is natur
 ally guaranteed.\nAfter the style customization, a shot design module is u
 sed to crop the scene image by the M-LLM for detailed shot design and to i
 ncrease the diversity of the story. Finally, for animation, given the moti
 on sequences from 3D proxy and the stylized prior, we then fine-tune the M
 MDiT-based image-to-video diffusion model to learn the complex motion of t
 he given foreground character. This is achieved by a motion customization 
 adapter with a timestep-shift strategy to keep long-term motion fidelity a
 nd coherence. After training, this model can be directly used on the cropp
 ed shots for generating diverse video scenes. Overall, we conduct extensiv
 e experiments and evaluations to demonstrate that FairyGen produces animat
 ions that are stylistically faithful, narratively aligned, and rich in nat
 ural, smooth motion, highlighting its effectiveness and flexibility for pe
 rsonalized story animation.\n\nRegistration Category: Full Access, Full Ac
 cess Supporter\n\nSession Chair: Wanchao Su (Monash University)\n\n
END:VEVENT
END:VCALENDAR