add ICRA paper

samwli · Apr 25, 2024 · ed95a59 · ed95a59
1 parent 8f696bd
commit ed95a59
Show file tree

Hide file tree

Showing 9 changed files with 21 additions and 9 deletions.
diff --git a/_posts/2023-12-16-affordance.markdown → _posts/2023-12-16-exploration.markdown b/_posts/2023-12-16-affordance.markdown → _posts/2023-12-16-exploration.markdown
@@ -7,6 +7,6 @@ categories: CMU
 course: "16-831 Intro to Robot Learning"
 author: "Samuel Li"
 authors: "<strong>Samuel Li*</strong>, Yunchao Yao*"
-paper: /pdfs/aff.pdf
+paper: /pdfs/efficient_exploration.pdf
 ---
 We leverage affordance understanding in foundation models for efficient, safe, and aligned task-conditioned exploration and learning for robotic manipulation.
diff --git a/_posts/2024-01-10-act_ant copy.markdown → _posts/2024-01-10-NeSCA.markdown b/_posts/2024-01-10-act_ant copy.markdown → _posts/2024-01-10-NeSCA.markdown
@@ -7,7 +7,7 @@ categories: research
 author: "Samuel Li"
 authors: "Sarthak Bhagat, <strong>Samuel Li</strong>, Joseph Campbell, Yaqi Xie, Katia Sycara, Simon Stepputtis"
 venue: "RA-L (In Review)"
-paper: /pdfs/act_ant.pdf
+paper: /pdfs/NeSCA.pdf
 website: https://sarthak268.github.io/NeSCA/
 ---
 We develop a novel modification to the transformer architecture for action anticipation, halving the observation time needed and achieving state-of-the-art performance to enable effective real-world human-robot collaboration.
diff --git a/_posts/2024-03-16-ShapeGrasp.markdown b/_posts/2024-03-16-ShapeGrasp.markdown
@@ -1,13 +1,12 @@
 ---
 layout: post
-title:  "ShapeGrasp: Zero-Shot Task-Oriented Grasping with Large Language Models through Geometric Decomposition"
-date:   2024-03-16 00:00:00 +00:00
-image: /images/ShapeGrasp.png
+title:  "Geometric Shape Reasoning for Zero-Shot Task-Oriented Grasping"
+date:   2024-04-25 00:00:00 +00:00
+image: /images/geometric_grasp.png
 categories: research
 author: "Samuel Li"
 authors: "<strong>Samuel Li</strong>, Sarthak Bhagat, Joseph Campbell, Yaqi Xie, Woojun Kim, Katia Sycara, Simon Stepputtis"
-venue: "IROS (In Review)"
-paper: /pdfs/ShapeGrasp.pdf
-website: https://shapegrasp.github.io/
+venue: "3D Visual Representations for Robot Manipulation Workshop, ICRA"
+paper: /pdfs/graspICRA.pdf
 ---
-We develop an efficient novel zero-shot, task-oriented object grasping pipeline using static, noisy monocular RGBD input and LLMs for fine-grained, shape-based part affordance reasoning.
+We propose a lightweight zero-shot, task-oriented grasping approach utilizing LLMs for fine-grained semantic reasoning over geometric decompositions. 
diff --git a/_posts/2024-04-25-graspICRA.markdown b/_posts/2024-04-25-graspICRA.markdown
@@ -0,0 +1,13 @@
+---
+layout: post
+title:  "ShapeGrasp: Zero-Shot Task-Oriented Grasping with Large Language Models through Geometric Decomposition"
+date:   2024-03-16 00:00:00 +00:00
+image: /images/ShapeGrasp.png
+categories: research
+author: "Samuel Li"
+authors: "<strong>Samuel Li</strong>, Sarthak Bhagat, Joseph Campbell, Yaqi Xie, Woojun Kim, Katia Sycara, Simon Stepputtis"
+venue: "IROS (In Review)"
+paper: /pdfs/ShapeGrasp.pdf
+website: https://shapegrasp.github.io/
+---
+We develop an efficient novel zero-shot, task-oriented object grasping pipeline using static, noisy monocular RGBD input and LLMs for fine-grained, shape-based part affordance reasoning.
diff --git a/pdfs/act_ant.pdf → pdfs/NeSCA.pdf b/pdfs/act_ant.pdf → pdfs/NeSCA.pdf
diff --git a/pdfs/aff.pdf → pdfs/efficient_exploration.pdf b/pdfs/aff.pdf → pdfs/efficient_exploration.pdf
diff --git a/pdfs/graspICRA.pdf b/pdfs/graspICRA.pdf
diff --git a/pdfs/samuel_li_resume.pdf b/pdfs/samuel_li_resume.pdf
diff --git a/tn/images/geometric_grasp.png b/tn/images/geometric_grasp.png