Weiyu Liu
@Weiyu_Liu_
Followers
1K
Following
347
Media
10
Statuses
66
Postdoc @Stanford. I work on semantic representations for robots. Previously PhD @GTrobotics
Joined October 2021
Congratulations to @neil_nie_ and the team on the launch! Neil has worked with us at Stanford over the past two years on learning generalizable long-horizon manipulation from as few as five demonstrations. I’m excited to see where his next adventure takes him and how our.
Thank you for sharing our work @ycombinator! Please checkout to schedule a free pilot to see how our robots can transform your business!. -.Thank you to (co-founder) Aditya, my research advisors, colleagues, friends and family for your support!.
0
1
14
RT @neil_nie_: Thank you for sharing our work @ycombinator! Please checkout to schedule a free pilot to see how our….
vernerobotics.com
Verne Robotics provides AI software for robotic arms with no cap-ex. Physical AI within reach. Train in hours and automate today.
0
12
0
RT @ManlingLi_: Today is the day! Welcome to join @CVPR workshop on Foundation Models meet Embodied Agents!. 🗓️Jun 11.📍Room 214.🌐 https://t.….
0
20
0
How to extract spatial knowledge from VLMs to generate 3D layouts?. We combine spatial relations, visual markers, and code into a unified representation that is both interpretable by VLMs and flexible for generating diverse scenes. Check out the detailed post by @sunfanyun!.
Spatial reasoning is a major challenge for the foundation models today, even in simple tasks like arranging objects in 3D space. #CVPR2025 .Introducing LayoutVLM, a differentiable optimization framework that uses VLM to spatially reason about diverse scene layouts from unlabeled
0
3
21
RT @YunfanJiang: 🤖 Ever wondered what robots need to truly help humans around the house?. 🏡 Introducing 𝗕𝗘𝗛𝗔𝗩𝗜𝗢𝗥 𝗥𝗼𝗯𝗼𝘁 𝗦𝘂𝗶𝘁𝗲 (𝗕𝗥𝗦)—a compre….
0
139
0
RT @keshigeyan: 1/ [NeurIPS D&B] Introducing HourVideo: A benchmark for hour-long video-language understanding!🚀. 500 egocentric videos, 18….
0
52
0
To build complex 3D structures with many parts, like IKEA furniture, we usually follow instructions and how-to videos. We introduce a new dataset with dense annotations on internet videos to study the grounding of video instructions in 3D. Check out the thread for details!.
💫🪑Introducing IKEA Manuals at Work: The first multimodal dataset with extensive 4D groundings of assembly in internet videos! We track furniture parts’ 6-DoF poses and segmentation masks through the assembly process, revealing how parts connect in both 2D and 3D space. With
0
4
27
What can we learn from demonstrations of long-horizon tasks? I am presenting our #CoRL2024 paper "Learning Compositional Behaviors from Demonstration and Language" today, showing we can learn a library of behaviors that can be composed to solve new tasks.
0
7
48
RT @zhang_yunzhi: Accurate and controllable scene generation has been difficult with natural language alone. You instead need a language f….
0
64
0
RT @RogerDai1217: Why hand-engineer digital twins when digital cousins are free?. Check out ACDC: Automated Creation of Digital Cousins 👭 f….
0
39
0
RT @stephentian_: Learned visuomotor robot policies are sensitive to observation viewpoint shifts, which happen all the time. Can visual pr….
0
19
0
RT @wenlong_huang: What structural task representation enables multi-stage, in-the-wild, bimanual, reactive manipulation?. Introducing ReKe….
0
105
0
RT @YunfanJiang: Does your sim2real robot falter at critical moments 🤯? Want to help but unsure how, all you can do is reward tuning in sim….
0
43
0
RT @IeeeTro: A T-RO paper by researchers @URoboticsCenter, @nvidia, @HRLLaboratories and @GeorgiaTech describes a #robot that can rearrange….
0
1
0
RT @chenwang_j: Can we use wearable devices to collect robot data without actual robots?. Yes! With a pair of gloves🧤!. Introducing DexCap,….
0
137
0
RT @KarmeshYadav: ICRA workshop dates are out. The first workshop on Vision-Language Models for Navigation and Manipulation will be held on….
openreview.net
Welcome to the OpenReview homepage for IEEE 2024 ICRA Workshop VLMNM
0
2
0
Can robots choose actions (position objects by pushing) and precisely execute them (decide the push trajectories) to fulfill abstract goals? . We use a relational model that considers actions and the resulting dynamics between objects and environments. Check out Yixuan's 🧵!.
Excited to share our TR-O paper about eRD-Bot, a robot that can rearrange novel objects in diverse environments with logical goals by flexibly combining primitive actions including pick, place and push. 🔗 🔗 🧵
1
3
27
🦾Excited to organize my first workshop focused on LLM/VLMs in Robotics, especially mobile manipulation. We planned for an engaging mix of presentations & debates about opportunities, challenges, and best practices. Don’t miss out on being part of this discussion!.
Large Language Models (LLMs) and Vision-Language Models (VLMs) are poised to revolutionize robotics. Join our workshop at #ICRA2024 on VLMs/LLMs for scene understanding, decision making, control, and more: Submissions due March 11, 2024!
1
4
45
Come check out our poster this afternoon to discuss about composable diffusion models and grounding spatial relations for generalizable manipulation 🤩.
Composable Part-based Manipulation uses object-part decomposition & part-part correspondences to improve generalization of robotic manipulation across object categories. @Weiyu_Liu_.Poster: Thu 9 2:45-3:30 pm.Website: w/ Tucker Hermans & @animesh_garg
0
6
40
RT @joycjhsu: What’s left w/ foundation models? We found that they still can't ground modular concepts across domains. We present Logic-En….
0
28
0