analiza aiqbench
This paper applies learned models to robotic cloth manipulation: a fine-tuned vision-language-action model generates simulated folding and flattening trajectories, which then train downstream RGB policies. For AI readers, the useful takeaways are the data-generation loop (filtering model rollouts by success criteria and re-rendering them under random appearances) and the reported gains over training on human demonstrations alone. The authors also note that generalization to unseen garments remains a shared weakness across policy models.
ClothDojo: Learned Data Generation and Benchmarking for Bimanual Garment Flattening and Folding
arXiv:2611.01190v1 Announce Type: new Abstract: Garment flattening and folding require robots to identify garment structure during manipulation and adapt their actions as the cloth deforms. We present ClothDojo, a simulation platform that generates adaptive manipulation trajectories at scale across diverse garments and initial states and evaluates policies in closed loop. ClothDojo encodes each garment's canonical mesh coordinates as an NOCS texture, providing canonical surface-position cues through the visual interface of a pretrained vision-language-action model. We fine-tune the model on human demonstrations and corrections to obtain a data-generation policy. The policy autonomously generates simulated manipulation trajectories, and we retain rollouts that satisfy task-specific success criteria. Each retained trajectory is re-rendered under a randomly sampled appearance to provide RGB observations for downstream policies that do not receive NOCS. Across both tasks, the policy produces nearly 8,000 successful trajectories over hundreds of garments. RGB policies trained on these generated trajectories outperform those trained directly on the original human demonstrations and corrections across all evaluated garment splits and appearance settings. Evaluations across multiple policy models show that overall performance and appearance robustness vary across models and tasks, while generalization to unseen garments remains a common challenge.
Abstract:Garment flattening and folding require robots to identify garment structure during manipulation and adapt their actions as the cloth deforms. We present ClothDojo, a simulation platform that generates adaptive manipulation trajectories at scale across diverse garments and initial states and evaluates policies in closed loop. ClothDojo encodes each garment's canonical mesh coordinates as an NOCS texture, providing canonical surface-position cues through the visual interface of a pretrained vision-language-action model. We fine-tune the model on human demonstrations and corrections to obtain a data-generation policy. The policy autonomously generates simulated manipulation trajectories, and we retain rollouts that satisfy task-specific success criteria. Each retained trajectory is re-rendered under a randomly sampled appearance to provide RGB observations for downstream policies that do not receive NOCS. Across both tasks, the policy produces nearly 8,000 successful trajectories over hundreds of garments. RGB policies trained on these generated trajectories outperform those trained directly on the original human demonstrations and corrections across all evaluated garment splits and appearance settings. Evaluations across multiple policy models show that overall performance and appearance robustness vary across models and tasks, while generalization to unseen garments remains a common challenge.
8 pages, 6 figures, 7 tables. Project page: this https URL
Robotics (cs.RO)
arXiv:2611.01190 [cs.RO]
(or arXiv:2611.01190v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2611.01190 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 291 cuvinte
Subiecte conexe
A Living Benchmark for Information Retrieval from Electronic Health Records
arXiv · 25 sept. 2026
An Evolutionary Framework for Automatic Optimization Benchmark Generation via Large Language Models
arXiv · 9 sept. 2026
DeepSurvey-Bench: Evaluating Academic Value of Automatically Generated Scientific Surveys
arXiv · 5 aug. 2026
MMGR: Multi-Modal Generative Reasoning Benchmark and Evaluation
arXiv · 14 sept. 2026