Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
110 tokens/sec
GPT-4o
56 tokens/sec
Gemini 2.5 Pro Pro
44 tokens/sec
o3 Pro
6 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Cross-Embodied Affordance Transfer through Learning Affordance Equivalences (2404.15648v2)

Published 24 Apr 2024 in cs.RO, cs.AI, and cs.LG

Abstract: Affordances represent the inherent effect and action possibilities that objects offer to the agents within a given context. From a theoretical viewpoint, affordances bridge the gap between effect and action, providing a functional understanding of the connections between the actions of an agent and its environment in terms of the effects it can cause. In this study, we propose a deep neural network model that unifies objects, actions, and effects into a single latent vector in a common latent space that we call the affordance space. Using the affordance space, our system can generate effect trajectories when action and object are given and can generate action trajectories when effect trajectories and objects are given. Our model does not learn the behavior of individual objects acted upon by a single agent. Still, rather, it forms a `shared affordance representation' spanning multiple agents and objects, which we call Affordance Equivalence. Affordance Equivalence facilitates not only action generalization over objects but also Cross Embodiment transfer linking actions of different robots. In addition to the simulation experiments that demonstrate the proposed model's range of capabilities, we also showcase that our model can be used for direct imitation in real-world settings.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (41)
  1. E. Şahin, M. Cakmak, M. R. Doğar, E. Uğur, and G. Üçoluk, “To afford or not to afford: A new formalization of affordances toward affordance-based robot control,” Adaptive Behavior, vol. 15, no. 4, pp. 447–472, 2007.
  2. B. D. Argall, S. Chernova, M. Veloso, and B. Browning, “A survey of robot learning from demonstration,” Robotics and autonomous systems, vol. 57, no. 5, pp. 469–483, 2009.
  3. T. Asfour, P. Azad, F. Gyarfas, and R. Dillmann, “Imitation learning of dual-arm manipulation tasks in humanoid robots,” International journal of humanoid robotics, vol. 5, no. 02, pp. 183–202, 2008.
  4. H. B. Amor, O. Kroemer, U. Hillenbrand, G. Neumann, and J. Peters, “Generalization of human grasping for multi-fingered robot hands,” in 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems.   IEEE, 2012, pp. 2043–2050.
  5. M. Mühlig, M. Gienger, and J. J. Steil, “Interactive imitation learning of object movement skills,” Autonomous Robots, vol. 32, pp. 97–114, 2012.
  6. A. Paraschos, C. Daniel, J. Peters, and G. Neumann, “Using probabilistic movement primitives in robotics,” Autonomous Robots, vol. 42, pp. 529–551, 2018.
  7. P. Pastor, H. Hoffmann, T. Asfour, and S. Schaal, “Learning and generalization of motor skills by learning from demonstration,” in 2009 IEEE International Conference on Robotics and Automation.   IEEE, 2009, pp. 763–768.
  8. P. Pastor, L. Righetti, M. Kalakrishnan, and S. Schaal, “Online movement adaptation based on previous sensor experiences,” in 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems.   IEEE, 2011, pp. 365–371.
  9. S. Calinon, “A tutorial on task-parameterized movement learning and retrieval,” Intelligent service robotics, vol. 9, pp. 1–29, 2016.
  10. Y. Huang, L. Rozo, J. Silvério, and D. G. Caldwell, “Kernelized movement primitives,” The International Journal of Robotics Research, vol. 38, no. 7, pp. 833–852, 2019.
  11. S. Schaal, “Dynamic movement primitives-a framework for motor control in humans and humanoid robotics,” in Adaptive motion of animals and machines.   Springer, 2006, pp. 261–280.
  12. Y. Zhou and T. Asfour, “Task-oriented generalization of dynamic movement primitive,” in 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).   IEEE, 2017, pp. 3202–3209.
  13. C. Atkeson, A. Moore, and S. Schaal, “Locally weighted learning for control. lazy learning,” 1997.
  14. A. Kramberger, A. Gams, B. Nemec, D. Chrysostomou, O. Madsen, and A. Ude, “Generalization of orientation trajectories and force-torque profiles for robotic assembly,” Robotics and autonomous systems, vol. 98, pp. 333–346, 2017.
  15. A. Ude, A. Gams, T. Asfour, and J. Morimoto, “Task-specific generalization of discrete and periodic dynamic movement primitives,” IEEE Transactions on Robotics, vol. 26, no. 5, pp. 800–815, 2010.
  16. S. Vijayakumar and S. Schaal, “Locally weighted projection regression: An o (n) algorithm for incremental real time learning in high dimensional space,” in Proceedings of the seventeenth international conference on machine learning (ICML 2000), vol. 1.   Morgan Kaufmann, 2000, pp. 288–293.
  17. S. Calinon, P. Evrard, E. Gribovskaya, A. Billard, and A. Kheddar, “Learning collaborative manipulation tasks by demonstration using a haptic interface,” in 2009 International Conference on Advanced Robotics.   IEEE, 2009, pp. 1–6.
  18. A. Pervez and D. Lee, “Learning task-parameterized dynamic movement primitives using mixture of gmms,” Intelligent Service Robotics, vol. 11, no. 1, pp. 61–78, 2018.
  19. V. Chu, I. McMahon, L. Riano, C. G. McDonald, Q. He, J. M. Perez-Tejada, M. Arrigo, N. Fitter, J. C. Nappo, T. Darrell et al., “Using robotic exploratory procedures to learn the meaning of haptic adjectives,” in 2013 IEEE International Conference on Robotics and Automation.   IEEE, 2013, pp. 3048–3055.
  20. H. Girgin and E. Ugur, “Associative skill memory models,” in 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).   IEEE, 2018, pp. 6043–6048.
  21. D. Lee and C. Ott, “Incremental kinesthetic teaching of motion primitives using the motion refinement tube,” Autonomous Robots, vol. 31, pp. 115–131, 2011.
  22. E. Ugur and H. Girgin, “Compliant parametric dynamic movement primitives,” Robotica, vol. 38, no. 3, pp. 457–474, 2020.
  23. A. Droniou, S. Ivaldi, and O. Sigaud, “Deep unsupervised network for multimodal perception, representation and classification,” Robotics and Autonomous Systems, vol. 71, pp. 83–98, 2015.
  24. A. Gams, A. Ude, J. Morimoto et al., “Deep encoder-decoder networks for mapping raw images to dynamic movement primitives,” in 2018 IEEE international conference on robotics and automation (ICRA).   IEEE, 2018, pp. 5863–5868.
  25. A. Pervez, Y. Mao, and D. Lee, “Learning deep movement primitives using convolutional neural networks,” in 2017 IEEE-RAS 17th international conference on humanoid robotics (Humanoids).   IEEE, 2017, pp. 191–197.
  26. F. Xie, A. Chowdhury, M. De Paolis Kaluza, L. Zhao, L. Wong, and R. Yu, “Deep imitation learning for bimanual robotic manipulation,” Advances in neural information processing systems, vol. 33, pp. 2327–2337, 2020.
  27. M. Y. Seker, M. Imre, J. H. Piater, and E. Ugur, “Conditional neural movement primitives.” in Robotics: Science and Systems, vol. 10, 2019.
  28. M. Garnelo, D. Rosenbaum, C. J. Maddison, T. Ramalho, D. Saxton, M. Shanahan, Y. W. Teh, D. J. Rezende, and S. Eslami, “Conditional neural processes,” arXiv preprint arXiv:1807.01613, 2018.
  29. M. E. Taylor and P. Stone, “Transfer learning for reinforcement learning domains: A survey.” Journal of Machine Learning Research, vol. 10, no. 7, 2009.
  30. Z. Zhu, K. Lin, A. K. Jain, and J. Zhou, “Transfer learning in deep reinforcement learning: A survey,” IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023.
  31. W. Yu, J. Tan, C. K. Liu, and G. Turk, “Preparing for the unknown: Learning a universal policy with online system identification,” arXiv preprint arXiv:1702.02453, 2017.
  32. T. Chen, A. Murali, and A. Gupta, “Hardware conditioned policies for multi-robot transfer learning,” Advances in Neural Information Processing Systems, vol. 31, 2018.
  33. J. Yang, D. Sadigh, and C. Finn, “Polybot: Training one policy across robots while embracing variability,” arXiv preprint arXiv:2307.03719, 2023.
  34. P. Christiano, Z. Shah, I. Mordatch, J. Schneider, T. Blackwell, J. Tobin, P. Abbeel, and W. Zaremba, “Transfer from simulation to real world through learning deep inverse dynamics model,” arXiv preprint arXiv:1610.03518, 2016.
  35. Q. Zhang, T. Xiao, A. A. Efros, L. Pinto, and X. Wang, “Learning cross-domain correspondence for control with dynamics cycle-consistency,” arXiv preprint arXiv:2012.09811, 2020.
  36. G. Konidaris and A. Barto, “Autonomous shaping: Knowledge transfer in reinforcement learning,” in Proceedings of the 23rd international conference on Machine learning, 2006, pp. 489–496.
  37. A. A. Rusu, M. Večerík, T. Rothörl, N. Heess, R. Pascanu, and R. Hadsell, “Sim-to-real robot learning from pixels with progressive nets,” in Conference on robot learning.   PMLR, 2017, pp. 262–270.
  38. X. Liu, D. Pathak, and D. Zhao, “Meta-evolve: Continuous robot evolution for one-to-many policy transfer,” in The Twelfth International Conference on Learning Representations, 2023.
  39. M. Akbulut, E. Oztop, M. Y. Seker, X. Hh, A. Tekden, and E. Ugur, “Acnmp: Skill transfer and task extrapolation through learning from demonstration and reinforcement learning via representation sharing,” in Conference on Robot Learning.   PMLR, 2021, pp. 1896–1907.
  40. H. Aktas, Y. Nagai, M. Asada, E. Oztop, and E. Ugur, “Correspondence learning between morphologically different robots via task demonstrations,” arXiv preprint arXiv:2310.13458, 2023.
  41. E. Rohmer, S. P. N. Singh, and M. Freese, “Coppeliasim (formerly v-rep): a versatile and scalable robot simulation framework,” in IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Tokyo, Japan, 2013.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (6)
  1. Hakan Aktas (6 papers)
  2. Yukie Nagai (12 papers)
  3. Minoru Asada (12 papers)
  4. Erhan Oztop (22 papers)
  5. Emre Ugur (37 papers)
  6. Matteo Saveriano (44 papers)

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now