1.0Situational Safety for Embodied Agentshttps://mscvprojects.ri.cmu.edu/2026teamf14rich600338<blockquote class="wp-embedded-content" data-secret="6bgoeab9lC"><a href="https://mscvprojects.ri.cmu.edu/2026teamf14/method/">Method</a></blockquote><iframe sandbox="allow-scripts" security="restricted" src="https://mscvprojects.ri.cmu.edu/2026teamf14/method/embed/#?secret=6bgoeab9lC" width="600" height="338" title="“Method” — Situational Safety for Embodied Agents" data-secret="6bgoeab9lC" frameborder="0" marginwidth="0" marginheight="0" scrolling="no" class="wp-embedded-content"></iframe><script> /*! This file is auto-generated */ !function(d,l){"use strict";l.querySelector&&d.addEventListener&&"undefined"!=typeof URL&&(d.wp=d.wp||{},d.wp.receiveEmbedMessage||(d.wp.receiveEmbedMessage=function(e){var t=e.data;if((t||t.secret||t.message||t.value)&&!/[^a-zA-Z0-9]/.test(t.secret)){for(var s,r,n,a=l.querySelectorAll('iframe[data-secret="'+t.secret+'"]'),o=l.querySelectorAll('blockquote[data-secret="'+t.secret+'"]'),c=new RegExp("^https?:$","i"),i=0;i<o.length;i++)o[i].style.display="none";for(i=0;i<a.length;i++)s=a[i],e.source===s.contentWindow&&(s.removeAttribute("style"),"height"===t.message?(1e3<(r=parseInt(t.value,10))?r=1e3:~~r<200&&(r=200),s.height=r):"link"===t.message&&(r=new URL(s.getAttribute("src")),n=new URL(t.value),c.test(n.protocol))&&n.host===r.host&&l.activeElement===s&&(d.top.location.href=t.value))}},d.addEventListener("message",d.wp.receiveEmbedMessage,!1),l.addEventListener("DOMContentLoaded",function(){for(var e,t,s=l.querySelectorAll("iframe.wp-embedded-content"),r=0;r<s.length;r++)(t=(e=s[r]).getAttribute("data-secret"))||(t=Math.random().toString(36).substring(2,12),e.src+="#?secret="+t,e.setAttribute("data-secret",t)),e.contentWindow.postMessage({message:"ready",secret:t},"*")},!1)))}(window,document); //# sourceURL=https://mscvprojects.ri.cmu.edu/2026teamf14/wp-includes/js/wp-embed.min.js </script> Overal pipeline Step 1: Semantic Constraint Extraction Our system first interprets the user’s task instruction together with multi-view RGB-D observations. A vision-language model identifies task-relevant objects, spatial relationships, and potential safety risks, converting open-ended semantic instructions into structured constraint predicates. Step 2: Geometric Grounding The extracted objects and relationships are grounded in 3D using depth […]http://www.contrib.andrew.cmu.edu/~jisuq/mscv/pipeline.png