Correll Lab
@correlllab.bsky.social
đ€ 336
đ„ 272
đ 94
Correll Lab at the University of Colorado Boulder.
Any opinions on "Toward Data Science" moving off of
Medium.com
? Wonder how many authors have published on TDS because it was on Medium.
loading . . .
Medium: Read and write stories.
On Medium, anyone can share insightful perspectives, useful knowledge, and life wisdom with the world.
https://Medium.com
about 1 year ago
0
0
0
Deep Seek just released the multi-modal model Janus-Pro
github.com/deepseek-ai/...
loading . . .
https://github.com/deepseek-ai/Janus/blob/main/janus_pro_tech_report.pdf
about 1 year ago
0
2
0
End-to-end learning datasets should include negative examples, check out James Watsonâs block-stacking dataset
medium.com/correll-l...
loading . . .
Dataset to Improve Performance in Open World Manipulation
The transformer architecture has changed the field of Artificial Intelligence forever in a sweeping fashion by bringing forth a newâŠ
https://medium.com/correll-lab/dataset-to-improve-performance-in-open-world-manipulation-7aa005e44003
about 1 year ago
0
2
0
We tested out-of-the-box VLM reasoning for robotic block stacking with ChatGPT4o (95%) a huge step up, but symbolic planning is still needed
medium.com/correll-l...
loading . . .
Embodied AI For Block Stacking Tasks
Embodied artificial intelligence is currently an active research frontier, foundation models like ChatGPT have demonstrated incredibleâŠ
https://medium.com/correll-lab/embodied-ai-for-block-stacking-tasks-01f0d8a8e667
about 1 year ago
0
0
0
Robot end-to-end learning gets better when including force
medium.com/correll-l...
loading . . .
Integrating Contact Force for Delicate Robot Grasping
Improving end-to-end learning with salient information
https://medium.com/correll-lab/integrating-contact-force-for-delicate-robot-grasping-babf68eccf08
about 1 year ago
0
2
1
More and more folks discover the power of tactile/force information in end-to-end learning manipulation
medium.com/correll-l...
loading . . .
A review on tactile end-to-end robot policies
Can learning unify the tactile experience across different sensing modalities?
https://medium.com/correll-lab/a-review-on-tactile-end-to-end-robot-policies-890f5c868576
about 1 year ago
0
1
1
Building CLIP from scratch to identify sport celebrities
medium.com/correll-l...
loading . . .
Building CLIP from Scratch to Identify Sports Celebrities
Unsupervised Learning Approach for Identifying Sports Legends in Images.
https://medium.com/correll-lab/building-clip-from-scratch-to-identify-sports-celebrities-7d0502edae3d
about 1 year ago
0
0
0
Check out Aritra and Himanshuâs attempt to solve BeliefMDPs with diffusion models
medium.com/correll-l...
loading . . .
DraftâââUsing Diffusion Models for BeliefMDPs
Robotics has come so far, and there are many examples of people achieving incredible tasks with themâââwhether itâs self-driving carsâŠ
https://medium.com/correll-lab/draft-using-diffusion-models-for-beliefmdps-e8374229d7ca
about 1 year ago
0
0
0
Two approaches of how to squeeze large (T)VLMs onto a 4090
medium.com/correll-l...
loading . . .
Bringing Tactile VLMs to Local GPUs
Scaling the size of the OCTOPI tactile VLM down to a more digestible size
https://medium.com/correll-lab/bringing-tactile-vlms-to-local-gpus-8f7ac082f306
about 1 year ago
0
0
0
Spectrogram and Vision Transformers a powerful tool for time-series analysis
medium.com/correll-l...
loading . . .
Spectrograms & Vision Transformers for Time Series Forecasting (Part 2)
An exploration of using multiple time series instead of one to improve financial time series forecasting.
https://medium.com/correll-lab/spectrograms-vision-transformers-for-time-series-forecasting-part-2-828733259da0
about 1 year ago
0
0
0
A tutorial on training multi-modal end-to-end robot policies with DROID
medium.com/correll-l...
loading . . .
Training Multimodal Robot Models with DROID
In this article Iâll be going through training a robot policy step-by-step with the DROID policy learning repository with RLDS format dataâŠ
https://medium.com/correll-lab/training-diffusion-policies-on-robot-grasps-with-force-feedback-35fbf7ec0dae
about 1 year ago
0
0
0
Using ControlNets to condition diffusion to in-paint 3D point clouds for manipulation
medium.com/correll-l...
loading . . .
Inpainting 3D Point Clouds Using Diffusion Networks
3D UNET Backbone With Adjusted Conditioning for Diffusion Networks on 3D Object Representation Generation
https://medium.com/correll-lab/inpainting-3d-point-clouds-using-diffusion-networks-f3aea7a58cea
about 1 year ago
0
0
0
Diffusion models are emerging as an universal frontend in robotic foundation models. Find out how they work in Yutong Zhangâs summary of the original robotic diffusion paper
medium.com/correll-l...
loading . . .
Diffusion Policy: Applying Diffusion Model to Policy Learning
Explaining diffusion policy, a powerful imitation learning method and the building block for robotics foundation models
https://medium.com/correll-lab/diffusion-policy-applying-diffusion-model-to-policy-learning-8ee16fc056bb
about 1 year ago
0
1
1
More and more folks discover the power of tactile/force information in end-to-end learning manipulation
medium.com/correll-l...
loading . . .
A review on tactile end-to-end robot policies
Can learning unify the tactile experience across different sensing modalities?
https://medium.com/correll-lab/a-review-on-tactile-end-to-end-robot-policies-890f5c868576
about 1 year ago
0
0
0
End-to-end learning datasets should include negative examples, check out James Watsonâs block-stacking dataset
medium.com/correll-l...
loading . . .
Dataset to Improve Performance in Open World Manipulation
The transformer architecture has changed the field of Artificial Intelligence forever in a sweeping fashion by bringing forth a newâŠ
https://medium.com/correll-lab/dataset-to-improve-performance-in-open-world-manipulation-7aa005e44003
about 1 year ago
0
2
0
reposted by
Correll Lab
Frank Dellaert
about 1 year ago
Just WOW:
youtu.be/X2UxtKLZnNo?...
loading . . .
Unitree B2-W Talent Awakening! đ„ł
YouTube video by Unitree Robotics
https://youtu.be/X2UxtKLZnNo?si=qjF9NVXimF0207sM
3
21
8
Robot end-to-end learning gets better when including force
medium.com/correll-l...
loading . . .
Integrating Contact Force for Delicate Robot Grasping
Improving end-to-end learning with salient information
https://medium.com/correll-lab/integrating-contact-force-for-delicate-robot-grasping-babf68eccf08
about 1 year ago
0
0
0
reposted by
Correll Lab
Codfather #FBPE đ«đ·đȘđș
about 1 year ago
đ
6
415
74
reposted by
Correll Lab
Remi Cadene
about 1 year ago
Open-source implementation of robot policy with Flow Matching đ„
github.com/HRI-EU/flow...
Great work from Fan Zhang and Michael Gienger from Honda Research Institute EU!
1
29
5
reposted by
Correll Lab
William Xie
about 1 year ago
i read five papers on tactile robot policies and wrote about them:
medium.com/correll-lab/...
the first step of me writing my prelim and i hope to have some more interesting synthesis soon. this is in no way a comprehensive review and i welcome recommendations of papers to read!
loading . . .
A review on tactile end-to-end robot policies
Can learning unify the tactile experience across different sensing modalities?
https://medium.com/correll-lab/a-review-on-tactile-end-to-end-robot-policies-890f5c868576
2
4
2
Check out Praneeth Nemaniâs work on using VLMs to provide scene understanding to symbolic planning
medium.com/correll-l...
loading . . .
Contextual Reasoning Meets Structured Planning
Modeling Problem Statements with LLMs
https://medium.com/correll-lab/contextual-reasoning-meets-structured-planning-modeling-problem-statements-with-llms-and-pddl-16a8364ab5af
about 1 year ago
0
2
0
Neural networks do historically bad on sets, check out Himanshu Guptaâs review on set transformers
medium.com/correll-l...
loading . . .
Paper Review: Set Transformers
A Framework for Attention-based Permutation-Invariant Neural Networks
https://medium.com/correll-lab/paper-review-set-transformers-6ce59c2ad88f
about 1 year ago
0
2
0
Follow and engage with our research on humanoid robots on
medium.com/correll-lab
loading . . .
Toward Humanoids
Ongoing pre-publication research on humanoids and related technologies including paper reviews and tutorials.
https://medium.com/correll-lab
about 1 year ago
0
1
0
We tested out-of-the-box VLM reasoning for robotic block stacking with ChatGPT4o (95%) a huge step up, but symbolic planning is still needed
medium.com/correll-l...
loading . . .
Embodied AI For Block Stacking Tasks
Embodied artificial intelligence is currently an active research frontier, foundation models like ChatGPT have demonstrated incredibleâŠ
https://medium.com/correll-lab/embodied-ai-for-block-stacking-tasks-01f0d8a8e667
about 1 year ago
0
0
0
Check out how ChatGPT can provide common sense knowledge for better grasping
medium.com/towards-d...
loading . . .
Grasping With Common Sense using VLMs and LLMs
How to leverage large language models for robotic grasping and code generation
https://medium.com/towards-data-science/grasping-with-common-sense-bfe21743c02d
about 1 year ago
0
1
1
Learn how DALL-E 2 works by building and training it from scratch
medium.com/correll-l...
loading . . .
DALL-E 2 from Scratch
Text-Conditioned Image Generation on FashionMNIST using CLIP Latents
https://medium.com/correll-lab/dall-e-2-from-scratch-c055bf881b9a
about 1 year ago
0
3
1
Here is how ChatGPT4o does on block stacking (spoiler: symbolic AI will not go away any time soon)
medium.com/correll-lab/...
loading . . .
Embodied AI For Block Stacking Tasks
Embodied artificial intelligence is currently an active research frontier, foundation models like ChatGPT have demonstrated incredibleâŠ
https://medium.com/correll-lab/embodied-ai-for-block-stacking-tasks-01f0d8a8e667
about 1 year ago
0
2
0
Interesting summary of the last three humanoid conferences by Aaron Prather
www.linkedin.com/posts/amprat...
loading . . .
Aaron Prather on LinkedIn: Over the past three months, I have attended A3 - Association for Advancing⊠| 28 comments
Over the past three months, I have attended A3 - Association for Advancing Automation đđźđŠđđ§đšđąđ đđšđđšđ đ đšđ«đźđŠ in Memphis, IEEE RAS HUMANOIDS⊠| 28 comments on LinkedIn
https://www.linkedin.com/posts/amprather_over-the-past-three-months-i-have-attended-activity-7274411808050892801-Ugb8?utm_source=share
about 1 year ago
2
0
0
Learn how to use pre-trained models like BERT and GPT2 to train CLIP on large models from scratch
medium.com/correll-l...
loading . . .
CLIP ModelâââAblation Study
In this ablation study, the âBuilding CLIP from Scratchâ by Matt Nguyen was leveraged. The reference source can be found hereâŠ
https://medium.com/correll-lab/clip-model-ablation-study-e52d22b5b9e9
about 1 year ago
0
2
0
I created a mini-tutorial on PDDL for those interested in exploring euro-symbolic AI
medium.com/towards-d...
loading . . .
Thinking, fast and slow with LLMs and PDDL
ChatGPT is never shy at pretending to perform deep thought, butâââlike our brainâââmight need better tools to reason accurately
https://medium.com/towards-data-science/thinking-fast-and-slow-with-llms-and-pddl-111699f9907e
about 1 year ago
1
4
0
Diffusion models can not only in-paint images, but also 3D point cloud. Xujia Zhangâs paper review on ScenceSense
medium.com/correll-l...
loading . . .
Paper review: SceneSense
Diffusion Models for 3D Occupancy Synthesis from Partial Observation
https://medium.com/correll-lab/paper-review-scenesense-a07fe204ba48
about 1 year ago
0
4
0
Lots of robotic applications for hierarchical vision transformers (SWIN)
medium.com/correll-l...
loading . . .
Paper Review: Swin Transformers and Robotics Applications
Autonomous driving, Geo-localization, Robotic surgery, and Robotic pose estimation
https://medium.com/correll-lab/paper-review-swin-transformers-and-robotics-applications-de60131fef96
about 1 year ago
0
2
1
Super-psyched to be able to work on a truly dirty-dull-dangerous application for humanoid robots starting next Spring
www.colorado.edu/cs/2024/12/0...
loading . . .
CS robotics research to help strengthen domestic battery supply chain
Nikolaus Correll Computer science professor Nikolaus Correll and his lab at CU Boulder have been awarded $1.8 million by the U.S. Department of Energy Advanced
https://www.colorado.edu/cs/2024/12/02/cs-robotics-research-help-strengthen-domestic-battery-supply-chain
about 1 year ago
0
1
0
How to train locally and merge the resulting weights, not the raw data itself, to learn using a team of robots, paper review by Lekai Chen
medium.com/correll-l...
loading . . .
Paper Review: Robot Fleet Learning via Policy Merging
How to train locally and merge the resulting weights, not the raw data itself, to learn using a team of robots
https://medium.com/correll-lab/paper-review-robot-fleet-learning-via-policy-merging-a1e8df483d02
about 1 year ago
0
1
0
Training VLMs on tactile data allows robots to understand and reason through touch, paper review by Carson Kohlbrenner
medium.com/correll-l...
loading . . .
Paper Review: Octopi: Object Property Reasoning with Large Tactile-Language Models
Enabling robots to understand and reason through touch!
https://medium.com/correll-lab/paper-review-octopi-object-property-reasoning-with-large-tactile-language-models-0f856dcdf8be
about 1 year ago
0
1
0
Vision language models are now able to resolve multi-step manipulation plans. Check out Stefan Caldararuâs summary of this recent CoRL paper
medium.com/correll-l...
loading . . .
Paper Review: Towards Open-World Grasping with Large Vision-Language Models
Using vision language models to resolve multi-step manipulation tasks
https://medium.com/correll-lab/paper-review-open-world-grasping-owg-7841fbc6e737
about 1 year ago
0
0
0
Is multi-tier online publication the way to break large publisherâs monopolies?
medium.com/correll-l...
loading . . .
Multi-tier online communication platformsâââa new paradigm for scientific publications?
The scientific publication system is kind of broken. On the one end of the spectrum are âhigh-impact publicationsâ that charge thousandsâŠ
https://medium.com/correll-lab/multi-tier-online-communication-platforms-a-new-paradigm-for-scientific-publications-cec980ba570c
about 1 year ago
0
0
0
Check out how ChatGPT can provide common sense knowledge for better grasping
medium.com/towards-d...
loading . . .
Grasping With Common Sense using VLMs and LLMs
How to leverage large language models for robotic grasping and code generation
https://medium.com/towards-data-science/grasping-with-common-sense-bfe21743c02d
about 1 year ago
0
0
0
I'm experimenting with a new (?) approach to science publication and communication that starts by putting an idea out on Bluesky to get initial feedback. So here it is! What do people think about this?
about 1 year ago
1
1
0
I created a mini-tutorial on PDDL for those interested in exploring euro-symbolic AI
medium.com/towards-d...
loading . . .
Thinking, fast and slow with LLMs and PDDL
ChatGPT is never shy at pretending to perform deep thought, butâââlike our brainâââmight need better tools to reason accurately
https://medium.com/towards-data-science/thinking-fast-and-slow-with-llms-and-pddl-111699f9907e
about 1 year ago
0
3
0
Learn how DALL-E 2 works by building and training it from scratch
medium.com/correll-l...
loading . . .
DALL-E 2 from Scratch
Text-Conditioned Image Generation on FashionMNIST using CLIP Latents
https://medium.com/correll-lab/dall-e-2-from-scratch-c055bf881b9a
about 1 year ago
0
4
2
Build CLIP from scratch using ResNet50 and BERT pre-trained embeddings (with code)
medium.com/correll-l...
loading . . .
CLIP implementation with pre-trained embeddings
Using ResNet50 and BERT to speed up training and increase accuracy
https://medium.com/correll-lab/clip-implementation-with-pre-trained-embeddings-9b2eca91dc22
about 1 year ago
0
0
0
Learn about Deep Transformer Q-Networks (and background) from Aritra Chakrabartyâs paper review
medium.com/correll-l...
loading . . .
Deep Transformer Q NetworksâââA paper analysis
The goal of this article is to analyze the paper âDeep Transformer Q-Networks for Partially Observable Reinforcement Learningâ by KevinâŠ
https://medium.com/correll-lab/deep-transformer-q-networks-a-paper-analysis-e7efd9379e5f
about 1 year ago
0
1
0
reposted by
Correll Lab
Embedded
about 1 year ago
This week,
@correlllab.bsky.social
spoke with Chris(
@stoneymonster.bsky.social
) and Elecia(
@logicalelegance.bsky.social
)about robots, teaching robotics, and writing books about robots. Join them here (
embedded.fm/episodes/490
). Here's an excerpt from the show:
#robotics
#robots
#embedded
0
4
2
Lots of robotic applications for hierarchical vision transformers (SWIN)
medium.com/correll-l...
about 1 year ago
0
0
0
About to order an Unitree H1-2. I expect 10% tariff. Did anyone order a Unitree robot recently and can confirm?
about 1 year ago
0
0
0
How to train locally and merge the resulting weights, not the raw data itself, to learn using a team of robots, paper review by Lekai Chen
medium.com/correll-l...
loading . . .
Paper Review: Robot Fleet Learning via Policy Merging
How to train locally and merge the resulting weights, not the raw data itself, to learn using a team of robots
https://medium.com/correll-lab/paper-review-robot-fleet-learning-via-policy-merging-a1e8df483d02
about 1 year ago
0
1
0
reposted by
Correll Lab
Maria Antoniak
over 1 year ago
I'm recruiting 1-2 PhD students to work with me at the University of Colorado Boulder! Looking for creative students with interests in
#NLP
and
#CulturalAnalytics
. Boulder is a lovely college town 30 minutes from Denver and 1 hour from Rocky Mountain National Park đ Apply by December 15th!
9
303
148
Training VLMs on tactile data allows robots to understand and reason through touch, paper review by Carson Kohlbrenner
medium.com/correll-l...
loading . . .
Paper Review: Octopi: Object Property Reasoning with Large Tactile-Language Models
Enabling robots to understand and reason through touch!
https://medium.com/correll-lab/paper-review-octopi-object-property-reasoning-with-large-tactile-language-models-0f856dcdf8be
over 1 year ago
0
1
1
Vision Transformers can analyze time series via their spectrogram, paper analysis and code by Peter Kinder
medium.com/correll-l...
loading . . .
Spectrograms & Vision Transformers for Time Series Forecasting
A review and implementation of J.P. Morgan AI Researchâs paper submission to the 2023 ACM International Conference on AI in Finance (ICAIF)
https://medium.com/correll-lab/spectrograms-vision-transformers-for-time-series-forecasting-929163fd8c52
over 1 year ago
0
0
0
Load more
feeds!
log in