@misc{indiciaea00418b7f89e,
  title = {VLMs Can Describe, But Not Measure: Object-Centric Scene Understanding for Robotic Manipulation},
  author = {Enrico Saccon and Tommaso Faraci and Iñigo De La Ossa Zarzuelo and Luigi Palopoli and Marco Roveri and Matteo Saveriano},
  year = {2026},
  url = {https://arxiv.org/abs/2609.28184},
  note = {Source identifier: 2609.28184}
}