목록으로
📄논문

ControlAnything: Generative UI and On-Demand Device Integration in Mixed Reality

Bektur Ryskeldiev
2026-11-05

소개

November 2026 preprint generating both spatial controls and the connector code for physical devices when a Quest 3 wearer requests them. A coding agent tests connections over hand-written MIDI, Bluetooth, and key-press transports, then returns AR widgets, a per-frame hand program, or both.

요약

Generating Controls and Their Connections:

1. Bound outputs: Generated widgets invoke actions declared by a running device connector
2. Local interaction: Hand programs execute every frame without further model calls; runtime checks constrain their outputs
3. Measured setup: Nine of ten requests reached ready in one author-operated session; median setup was 32.5 seconds, excluding speech transcription
4. Research boundary: Readiness establishes transport acknowledgements, not measured physical effects; targets used prior device knowledge and prepared transports, and laptop connectors are not sandboxed

[Author's code, emulator, and recorded demonstrations](https://github.com/Apolotary/controlanything).

태그

generative-uimixed-realitydevice-integrationhand-trackingruntime-ui

관련 글

📄논문

Interface Quilt: Modifying Applications with Computer Use Agents and Generative UI

November 2026 preprint from Autodesk Research combining computer-use agents and GenUI to build persistent interface patches over existing applications. It recovers controls from rendered pixels, generates new interaction surfaces, and synchronizes them with the underlying app without requiring source code, APIs, or accessibility data.

generative-uicomputer-use-agentsinterface-patches
작성자 Qiu L. et al. · 2026-11-01
자세히 보기
📄논문

ABRA: Agent-Bound UI for Human-Agent Communication beyond Chat with Reusable Views

재사용 가능한 그래프·표·텍스트·표시 뷰와 작업별 상호작용 의미를 분리하는 2026년 10월 EMNLP 시스템 데모입니다. ABRA는 이러한 바인딩을 공유 캔버스 상태로 공개하여 사용자와 다른 에이전트가 작업별 동작을 확인하고 호출할 수 있게 합니다.

generative-uishared-canvasesinteraction-semantics
작성자 Kim S. G. and Kim H. · 2026-10-22
자세히 보기
📄논문

Inflak: A Unified Orchestration Architecture for Human-Agent Interaction

런타임의 사람·에이전트 상호작용을 라우팅, 계획, 디자인, 렌더링으로 분리하는 2026년 10월 논문입니다. Inflak은 각 계층을 조합 가능한 에이전트 스킬로 구현하여, 작업 상태에 따라 입력을 요청할 시점과 표시할 생성형 인터페이스를 결정합니다.

generative-uiinteraction-orchestrationagent-skills
작성자 Gao L. et al. · 2026-10-15
자세히 보기

Generative UI와 Intelligent UI 리소스를 엄선한 모음입니다.

정성을 담아 ❤️ 커뮤니티가 만들었습니다