AI Daybook
AI Daybook
오늘의 지면
논문연구·논문

WorldDraft, 프로그래매틱 4D 시뮬레이션과 영상 생성 연결

WorldDraft: If Code Builds the World, Can Video Models Just Render It?

매체
Hugging Face
게시
중요도
4/5
화제 지수
13

한국어 요약

멀티모달 코딩 에이전트가 4D 세계 구성을 시각 생성과 분리하는 가운데, 영상 모델이 이를 렌더링할 수 있는지 검증하는 벤치마크를 제안했다. 6개 동적 도전 범주를 아우르는 432개 테스트 케이스로 12개 모델을 평가한 결과 세계 보존과 시각 실현 간 긴장, 소스 누출 등 병목을 확인했고, 시각 프롬프팅 전략으로 개선 방안을 제시한다.

원문 발췌 · English

Multimodal coding agents decouple 4D world construction from visual generation, repurposing video models from world generators toward generative renderers. This division of labor raises a fundamental question: if code builds the world, can video models just render it? A reliable generative renderer hinges on two complementary criteria: (i) world preservation, retaining prescribed 4D kinematics and camera paths; and (ii) visual realization, persistently binding text-specified target attributes over time. To evaluate this capability, we introduce WorldDraft, a benchmark of 432 strictly paired test cases across 48 executable 4D world families spanning six dynamic challenge categories. Under a 3 times 3 factorial design, WorldDraft orthogonally crosses draft granularity (abstract proxy to previs) and appearance alignment (neutral to conflicting) under identical dynamics. Extensive evaluations across 12 representative models reveal that: (I) world preservation and visual realization remain in stark operational tension; (II) richer draft scaffolding anchors structure but paradoxically impairs appearance fidelity; and (III) conflicting source cues trigger severe source leakage (up to 64%), exposing fragile cross-modal disentanglement. To mitigate these bottlenecks, we propose WorldDraft-Primer, a plug-and-play visual prompting strategy that consistently improves both objectives across frozen video renderers. Together, WorldDraft and WorldDraft-Primer lay a foundation for bridging programmatic 4D simulation with video generation, advancing physically grounded generative rendering.

#multimodal#benchmark#world model#video generation
원문 기사 읽기
전체 보기 →