Google's AgentHands adds spatially grounded, synchronized hand gestures to XR AI agents
CurieuxExplorer · x · 2026-10-13
Google Research unveiled AgentHands at CHI 2026, an LLM-powered XR prototype that augments conversational agents with expressive, voice-synchronized hand gestures for spatially grounded guidance.
- Gesture library and handedness: one or two hands, forms like a "palm" for caution or a "cylindrical grip" mimicking holding a tool
- Spatiality: gestures can float mid-air, anchor to object parts, or serve as user-relative social cues using XR depth
- Temporal dynamics and VFX: animated motions like "pouring" or "tracing", plus visual effects such as a red glow for heat warnings
- Positioned as an evolution beyond Project Astra and Gemini 3.1 Flash Live's 2D bounding-box overlays, turning abstract verbal instructions into intuitive embodied guidance for physical tasks
More from Embodied
- JAKA unveils Kargo50, a 50 kg heavy-duty wheeled humanoid already working in factories — CyberRobooo · 2026-10-13
- 30,000 hours of egocentric video: world model learns hands, not object outcomes — xuanalogue · 2026-10-13
- High schoolers build an autonomous ball-collecting robot that shoots on the move — lukas_m_ziegler · 2026-10-13
- NVIDIA RTX Spark launches Oct 16 with 128GB unified memory and 1 petaflop of local AI compute — nordicinst · 2026-10-13
- NVIDIA RTX Spark launches Oct 16: 128GB unified memory, 1 petaflop, Nemotron 3.5 Bolt 74B — NVIDIA Blog · 2026-10-13
- Ex-Apple engineer eyeing Surface with NVIDIA RTX Spark: 128GB + CUDA in a backpack — LinusEkenstam · 2026-10-13