Phase-Conditioned Imitation Learning with Autonomous Failure Recovery for Robust Deformable Object Manipulation

Dayuan Chen; Kai Tang; Kazuhiro Kosuge; Yasuhisa Hirata; Yukuan Zhang

arxiv: 2605.29407 · v1 · pith:AVISAYCZnew · submitted 2026-05-28 · 💻 cs.RO

Phase-Conditioned Imitation Learning with Autonomous Failure Recovery for Robust Deformable Object Manipulation

Dayuan Chen , Kai Tang , Yukuan Zhang , Kazuhiro Kosuge , Yasuhisa Hirata This is my paper

Pith reviewed 2026-06-29 07:11 UTC · model grok-4.3

classification 💻 cs.RO

keywords imitation learningdeformable object manipulationphase conditioningfailure recoverydual-arm roboticsforce feedbackACT policy

0 comments

The pith

Phase-conditioned imitation learning with multi-modal feedback enables autonomous recovery from failures in deformable object tasks, raising hanging success from 56% to 87%.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces a closed-loop hierarchical architecture that conditions an imitation learning policy on task phase to handle state aliasing in deformable object manipulation. A FiLM-conditioned ACT encoder produces phase-specific behaviors from a single unified policy while a multi-modal predictor fuses visual, force, and pose signals to estimate phase and detect contact failures invisible to vision. This setup triggers recovery trajectories autonomously and uses hybrid impedance control for compliant execution. Validation on dual-arm T-shirt hanging and removal demonstrates the improvement over standard Markovian policies. Ablation studies confirm that FiLM modulation creates separated feature representations for different phases.

Core claim

The authors claim that conditioning an Action Chunking with Transformers policy on estimated task phase via FiLM, combined with real-time multi-modal phase prediction from vision, force, and pose, produces a single policy that executes phase-specific actions and recovers autonomously from execution failures that vision alone cannot detect.

What carries the argument

FiLM-conditioned ACT encoder modulated by a multi-modal phase predictor that fuses visual, force, and pose feedback to estimate current task phase and trigger recovery.

If this is right

A single policy can produce phase-specific behaviors while sharing action dynamics across phases.
Contact failures invisible to vision can be detected and trigger autonomous recovery trajectories.
FiLM modulation yields well-separated phase-specific feature representations confirmed by t-SNE.
Hybrid impedance control supports compliant execution during both normal and recovery actions.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The approach could reduce the need for separate policies per phase in other multi-stage robotic tasks.
Force and pose feedback may help in environments where visual occlusion is common.
Similar phase conditioning might address state aliasing in other imitation learning methods beyond ACT.

Load-bearing premise

The multi-modal phase predictor can accurately estimate the current task phase in real time and reliably detect contact failures that are invisible to vision alone.

What would settle it

An experiment that disables or degrades the multi-modal phase predictor while keeping the rest of the system fixed and measures whether the hanging success rate falls back near 56%.

Figures

Figures reproduced from arXiv: 2605.29407 by Dayuan Chen, Kai Tang, Kazuhiro Kosuge, Yasuhisa Hirata, Yukuan Zhang.

**Figure 2.** Figure 2: Hierarchical Control Architecture. The system is organized into three levels: (Top) The decision layer uses multi-modal feedback to determine the [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 4.** Figure 4: The system overview with haptic feedback teleoperation. [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗

**Figure 5.** Figure 5: Upper: Sequence of the nominal hanging task. The process is decomposed into: (Phase 0-1) Collaborative insertion of the left collar; (Phase 2) [PITH_FULL_IMAGE:figures/full_fig_p005_5.png] view at source ↗

**Figure 6.** Figure 6: Recovery from insertion snagging (Phase 8). (a) Visual serials: Upon detecting a snag, the policy triggers a shaking motion to insert the collar. (b) [PITH_FULL_IMAGE:figures/full_fig_p005_6.png] view at source ↗

**Figure 7.** Figure 7: Upper: Recovery from gross misalignment (Phase 9). When the garment misses the hanger entirely (front/back), the policy executes a retreat maneuver [PITH_FULL_IMAGE:figures/full_fig_p006_7.png] view at source ↗

**Figure 8.** Figure 8: t-SNE visualization of ACT encoder outputs under varying phase [PITH_FULL_IMAGE:figures/full_fig_p007_8.png] view at source ↗

read the original abstract

This paper presents a phase-conditioned, force-aware framework for robust deformable object manipulation. Standard imitation learning policies such as Action Chunking with Transformers (ACT) rely on a Markovian assumption at inference, causing state aliasing when visually similar observations require contradictory actions and preventing autonomous recovery from execution failures. We address this with a closed-loop hierarchical architecture. A FiLM-conditioned ACT encoder modulates feature extraction based on the current task phase, enabling a single unified policy to produce phase-specific behaviors while sharing action dynamics across phases. A multi-modal phase predictor fusing visual, force, and pose feedback estimates the phase in real time, detecting contact failures that are invisible to vision alone and autonomously triggering recovery trajectories. The system is completed by a hybrid impedance controller for compliant execution and a haptic teleoperation interface for force-aware data collection. Ablation studies show that FiLM-based modulation significantly outperforms both unconditioned and token-level conditioned baselines, and t-SNE analysis confirms that FiLM induces well-separated, phase-specific feature representations. Validated on hanging and removing a T-shirt with dual arms, the closed-loop system improves the hanging success rate from 56\% to 87\% through autonomous error recovery. Code and videos: https://leledeyuan00.github.io/phaser/

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The FiLM-conditioned ACT plus multi-modal phase predictor gives a clear 31-point lift on dual-arm T-shirt hanging, but the gain is hard to attribute without any accuracy numbers on the predictor itself.

read the letter

The core contribution is a hierarchical setup that conditions an ACT policy on task phase via FiLM and adds a vision-force-pose predictor to spot failures and trigger recovery. This directly tackles the Markovian limitation in standard imitation learning for deformable objects where similar visuals can demand different actions.

What works is the ablation evidence that FiLM modulation beats both the plain ACT baseline and token-level conditioning, plus the t-SNE plots showing cleaner phase separation in feature space. The physical experiments on hanging and removing a T-shirt with two arms are the right kind of test, and the jump from 56% to 87% success is large enough to notice.

The soft spot is exactly the one in the stress-test note. The abstract credits the autonomous recovery for the improvement and claims the predictor catches contact failures invisible to vision, yet it gives no phase-classification accuracy, precision-recall, or confusion matrix on held-out data. Without those numbers it is impossible to know whether the closed-loop part is actually doing the work or whether the hybrid impedance controller and the conditioned policy alone would have produced most of the gain. The t-SNE and ablation results address feature quality but not real-time failure detection reliability under the exact task conditions.

This is for groups working on imitation learning for manipulation, especially anyone dealing with deformable objects or contact-rich tasks. The idea is practical and the hardware validation is there, so it deserves a serious referee even though the predictor validation needs to be strengthened before the central claim can be taken at face value.

Referee Report

2 major / 0 minor

Summary. The manuscript proposes a closed-loop hierarchical framework for deformable object manipulation that augments Action Chunking with Transformers (ACT) via FiLM conditioning on task phase, a multi-modal phase predictor fusing vision/force/pose to detect failures invisible to vision, a hybrid impedance controller, and haptic teleoperation. It claims this mitigates Markovian state aliasing in standard imitation learning and enables autonomous recovery, with the central empirical result being an increase in dual-arm T-shirt hanging success from 56% to 87%.

Significance. If the success-rate gain can be rigorously attributed to the phase predictor's failure detection rather than the FiLM policy or impedance controller, the work would provide a concrete mechanism for closed-loop recovery in contact-rich deformable manipulation without requiring explicit failure-mode engineering.

major comments (2)

[Abstract] Abstract: The 56% to 87% hanging success improvement is explicitly attributed to autonomous recovery triggered by the multi-modal phase predictor detecting contact failures invisible to vision, yet the manuscript supplies no phase-classification accuracy, confusion matrix, precision/recall, or real-time failure-detection metrics on held-out sequences or during the dual-arm task. Without these, it is impossible to confirm that the predictor (rather than FiLM conditioning or the hybrid controller) drives the reported gain.
[Ablation studies / t-SNE analysis] Ablation studies and t-SNE analysis: These demonstrate that FiLM induces better-separated phase-specific features than unconditioned or token-level baselines, but they do not isolate or quantify the phase predictor's contribution to failure recovery under the exact conditions of the T-shirt hanging task.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for their constructive feedback. We respond to each major comment below, indicating where the manuscript will be revised for clarification while maintaining the integrity of the reported results.

read point-by-point responses

Referee: [Abstract] Abstract: The 56% to 87% hanging success improvement is explicitly attributed to autonomous recovery triggered by the multi-modal phase predictor detecting contact failures invisible to vision, yet the manuscript supplies no phase-classification accuracy, confusion matrix, precision/recall, or real-time failure-detection metrics on held-out sequences or during the dual-arm task. Without these, it is impossible to confirm that the predictor (rather than FiLM conditioning or the hybrid controller) drives the reported gain.

Authors: We acknowledge that the manuscript does not include direct quantitative metrics (accuracy, confusion matrix, precision/recall) for the phase predictor on held-out data. The 56% baseline reflects the standard ACT policy without phase conditioning or the multi-modal predictor, while the 87% result is achieved only with the full closed-loop system that uses the predictor to detect vision-invisible contact failures and trigger recovery. The multi-modal fusion (vision/force/pose) is specifically motivated by the need to resolve state aliasing in contact-rich phases. In revision we will rephrase the abstract to attribute the gain to the integrated closed-loop framework rather than the predictor in isolation, and we will add a limitations paragraph noting the absence of isolated predictor metrics. revision: partial
Referee: [Ablation studies / t-SNE analysis] Ablation studies and t-SNE analysis: These demonstrate that FiLM induces better-separated phase-specific features than unconditioned or token-level baselines, but they do not isolate or quantify the phase predictor's contribution to failure recovery under the exact conditions of the T-shirt hanging task.

Authors: The ablation studies and t-SNE visualizations evaluate the benefit of FiLM conditioning on policy feature separation and phase-specific behavior. The predictor's contribution to failure recovery is evidenced indirectly through the end-to-end task results: the 31% absolute improvement occurs only when the multi-modal predictor is active and able to trigger recovery trajectories. No separate ablation that disables only the predictor (while retaining FiLM and the impedance controller) was performed. In the revised manuscript we will add an explicit discussion clarifying that the ablations target the policy encoder while the predictor's role is demonstrated via the recovery mechanism in the full dual-arm task. revision: partial

standing simulated objections not resolved

Phase-classification accuracy, confusion matrix, precision/recall, or real-time failure-detection metrics for the multi-modal phase predictor on held-out sequences

Circularity Check

0 steps flagged

No circularity: empirical performance metrics from physical experiments.

full rationale

The paper describes a hierarchical control architecture (FiLM-conditioned ACT, multi-modal phase predictor, hybrid impedance controller) and reports measured success rates (56% to 87%) on physical dual-arm T-shirt tasks. No derivation chain, equations, or first-principles predictions are present that could reduce to fitted parameters or self-referential definitions. Ablations and t-SNE visualizations are post-hoc analyses of learned representations, not load-bearing derivations. Self-citation is absent from the provided text; external baselines (ACT) are cited as standard methods. The central claim is an observed performance delta under closed-loop execution, which is falsifiable via experiment and does not collapse by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

No free parameters, invented entities, or non-standard axioms are introduced in the abstract; the work rests on standard domain assumptions in imitation learning and multi-modal robotics.

axioms (1)

domain assumption Standard imitation learning policies rely on a Markovian assumption that leads to state aliasing in phase-dependent tasks.
Explicitly stated as the motivation for adding phase conditioning.

pith-pipeline@v0.9.1-grok · 5769 in / 1179 out tokens · 36902 ms · 2026-06-29T07:11:44.936036+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

25 extracted references · 4 canonical work pages · 2 internal anchors

[1]

Multi-critic reinforcement learning for garment handling: Addressing unpredictability in temporal-phase con- tinuous contact tasks,

Y . Zhang, D. Chen, W. He, A. El ´ıas Petrilli Barcel ´o, J. Victorio Salazar Luces, and Y . Hirata, “Multi-critic reinforcement learning for garment handling: Addressing unpredictability in temporal-phase con- tinuous contact tasks,”IEEE Trans. on Automat. Sci. and Eng., vol. 22, pp. 10 741–10 752, 2025

2025
[2]

Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware,

T. Z. Zhao, V . Kumar, S. Levine, and C. Finn, “Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware,” inProc. of Robot.: Sci. and Syst., Daegu, Republic of Korea, July 2023

2023
[3]

Diffusion policy: Visuomotor policy learning via action diffusion,

C. Chi, Z. Xu, S. Feng, E. Cousineau, Y . Du, B. Burchfiel, R. Tedrake, and S. Song, “Diffusion policy: Visuomotor policy learning via action diffusion,”The Int. J. of Robot. Res., 2024

2024
[4]

RT-1: Robot. Transformer for Real-World Control at Scale,

A. Brohan, N. Brown, and J. C. et al., “RT-1: Robot. Transformer for Real-World Control at Scale,” inProc. of Robot.: Sci. and Syst., Daegu, Republic of Korea, July 2023

2023
[5]

Dasari, O

S. Dasari, O. Mees, S. Zhao, M. K. Srirama, and S. Levine, “The ingredients for robotic diffusion transformers,”arXiv preprint arXiv:2410.10088, 2024

work page arXiv 2024
[6]

Film: visual reasoning with a general conditioning layer,

E. Perez, F. Strub, H. de Vries, V . Dumoulin, and A. Courville, “Film: visual reasoning with a general conditioning layer,” inProceedings of the Thirty-Second AAAI Conf. on Artificial Intelligence and Thirtieth Innovative Applications of Artificial Intelligence Conf. and Eighth AAAI Symposium on Educational Advances in Artificial Intelligence, ser. AAAI’...

2018
[7]

Scil: Stage- Conditioned Imitation Learning for Multi-Stage Manipulation,

Z. Ouyang, K. Wang, J. Liu, H. Lu, and W. Zhang, “Scil: Stage- Conditioned Imitation Learning for Multi-Stage Manipulation,”IEEE Control Syst. Letters, vol. 9, pp. 913–918, 2025

2025
[8]

An incremental hybrid impedance algorithm for stable force-position control of elastic materials,

Y . Zhang, D. Chen, W. He, A. E. P. Barcel´o, J. V . S. Luces, and Y . Hirata, “An incremental hybrid impedance algorithm for stable force-position control of elastic materials,”IEEE Trans. on Automat. Sci. and Eng., pp. 1–1, 2025

2025
[9]

Goal-conditioned model simpli- fication for 1-d and 2-d deformable object manipulation,

S. Wang, M. Leonetti, and M. Dogar, “Goal-conditioned model simpli- fication for 1-d and 2-d deformable object manipulation,”IEEE Trans. on Robot., vol. 41, pp. 4023–4040, 2025

2025
[10]

Sis: Seam-informed strategy for t-shirt unfolding,

X. Huang, A. Seino, F. Tokuda, A. Kobayashi, D. Chen, Y . Hirata, N. C. Tien, and K. Kosuge, “Sis: Seam-informed strategy for t-shirt unfolding,” IEEE Robot. and Automat. Letters, vol. 10, no. 7, pp. 7342–7349, 2025

2025
[11]

RTFF: Random-to-Target Fabric Flattening Policy using Dual-Arm Manipulator

K. Tang, D. Bhattacharya, H. Xu, F. Tokuda, N. C. Tien, and K. Ko- suge, “Rtff: Random-to-target fabric flattening policy using dual-arm manipulator,” https://arxiv.org/abs/2510.00814, 2025

work page internal anchor Pith review Pith/arXiv arXiv 2025
[12]

Graph- garment: Learning garment dynamics for bimanual cloth manipulation tasks,

W. Chen, K. Li, D. Lee, X. Chen, R. Zong, and P. Kormushev, “Graph- garment: Learning garment dynamics for bimanual cloth manipulation tasks,” https://arxiv.org/abs/2503.05817, 2025

work page arXiv 2025
[13]

Bimanual deformable bag manipulation using a structure-of-interest based neural dynamics model,

P. Zhou, P. Zheng, J. Qi, C. Li, H.-Y . Lee, Y . Pan, C. Yang, D. Navarro- Alarcon, and J. Pan, “Bimanual deformable bag manipulation using a structure-of-interest based neural dynamics model,”IEEE/ASME Trans. on Mechatronics, vol. 30, no. 5, pp. 3254–3265, 2025

2025
[14]

Deformpam: Data- Efficient Learning for Long-Horizon Deformable Object Manipulation Via Preference-Based Action Alignment,

W. Chen, H. Xue, F. Zhou, Y . Fang, and C. Lu, “Deformpam: Data- Efficient Learning for Long-Horizon Deformable Object Manipulation Via Preference-Based Action Alignment,” in2025 IEEE Int. Conf. on Robot. and Automat. (ICRA). IEEE, may 19 2025, pp. 6896–6903

2025
[15]

Learning human-inspired force strategies for robotic assembly,

S. Scherzinger, A. Roennau, and R. Dillmann, “Learning human-inspired force strategies for robotic assembly,” in2023 IEEE 19th Int. Conf. on Automat. Sci. and Eng. (CASE), 2023, pp. 1–8

2023
[16]

A system for imitation learning of contact-rich bimanual manipulation policies,

S. Stepputtis, M. Bandari, S. Schaal, and H. B. Amor, “A system for imitation learning of contact-rich bimanual manipulation policies,” in 2022 IEEE/RSJ Int. Conf. on Itell. Robots and Syst. (IROS), 2022, pp. 11 810–11 817

2022
[17]

FILIC: Dual-Loop Force-Guided Imitation Learning with Impedance Torque Control for Contact-Rich Manipulation Tasks

H. Ge, Y . Jia, Z. Li, Y . Li, Z. Chen, R. Huang, and G. Zhou, “Filic: Dual- loop force-guided imitation learning with impedance torque control for contact-rich manipulation tasks,” https://arxiv.org/abs/2509.17053, 2025

work page internal anchor Pith review Pith/arXiv arXiv 2025
[18]

A reduction of imitation learning and structured prediction to no-regret online learning,

S. Ross, G. Gordon, and D. Bagnell, “A reduction of imitation learning and structured prediction to no-regret online learning,” inProceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics, ser. Proceedings of Machine Learning Research, G. Gordon, D. Dunson, and M. Dud ´ık, Eds., vol. 15. Fort Lauderdale, FL, USA: PML...

2011
[19]

Rac: Robot learning for long-horizon tasks by scaling recovery and correction,

Z. Hu, R. Wu, N. Enock, J. Li, R. Kadakia, Z. Erickson, and A. Kumar, “Rac: Robot learning for long-horizon tasks by scaling recovery and correction,” 2025

2025
[20]

On the continuity of rotation representations in neural networks,

Y . Zhou, C. Barnes, J. Lu, J. Yang, and H. Li, “On the continuity of rotation representations in neural networks,” https://arxiv.org/abs/1812. 07035, 2020

2020
[21]

Simplified offset and gravity compensation for wrist- mounted force/torque sensor with a garment manipulation application,

D. Chen, Y . Zhang, W. He, A. E. Petrilli Barcel ´o, J. V . Salazar Luces, and Y . Hirata, “Simplified offset and gravity compensation for wrist- mounted force/torque sensor with a garment manipulation application,” in2025 IEEE Int. Conf. on Robot. and Biomimetics (ROBIO), 2025

2025
[22]

Forward dynamics compliance control (fdcc): A new approach to cartesian compliance for robotic manipulators,

S. Scherzinger, A. Roennau, and R. Dillmann, “Forward dynamics compliance control (fdcc): A new approach to cartesian compliance for robotic manipulators,” inIEEE/RSJ Int. Conf. on Intell. Robots and Syst. (IROS), 2017, pp. 4568–4575

2017
[23]

Oculus reader: Robotic teleoperation interface,

F. E. Jedrzej Orbik, “Oculus reader: Robotic teleoperation interface,” https://github.com/rail-berkeley/oculus reader, 2021

2021
[24]

Coordinated motion control of dual manipulators for handling a rigid object with non-negligible deformation,

K. Kosuge, K. Kamei, and T. Nammoto, “Coordinated motion control of dual manipulators for handling a rigid object with non-negligible deformation,” in2014 IEEE Int. Conf. on Robot. and Automat. (ICRA), 2014, pp. 5145–5151

2014
[25]

Lerobot: State-of-the-art machine learning for real-world robotics in pytorch,

R. Cadene, S. Alibert, A. Soare, Q. Gallouedec, A. Zouitine, S. Palma, P. Kooijmans, M. Aractingi, M. Shukor, D. Aubakirova, M. Russi, F. Capuano, C. Pascal, J. Choghari, J. Moss, and T. Wolf, “Lerobot: State-of-the-art machine learning for real-world robotics in pytorch,” https://github.com/huggingface/lerobot, 2024. 9 Dayuan Chen(Graduate Student Member...

2024

[1] [1]

Multi-critic reinforcement learning for garment handling: Addressing unpredictability in temporal-phase con- tinuous contact tasks,

Y . Zhang, D. Chen, W. He, A. El ´ıas Petrilli Barcel ´o, J. Victorio Salazar Luces, and Y . Hirata, “Multi-critic reinforcement learning for garment handling: Addressing unpredictability in temporal-phase con- tinuous contact tasks,”IEEE Trans. on Automat. Sci. and Eng., vol. 22, pp. 10 741–10 752, 2025

2025

[2] [2]

Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware,

T. Z. Zhao, V . Kumar, S. Levine, and C. Finn, “Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware,” inProc. of Robot.: Sci. and Syst., Daegu, Republic of Korea, July 2023

2023

[3] [3]

Diffusion policy: Visuomotor policy learning via action diffusion,

C. Chi, Z. Xu, S. Feng, E. Cousineau, Y . Du, B. Burchfiel, R. Tedrake, and S. Song, “Diffusion policy: Visuomotor policy learning via action diffusion,”The Int. J. of Robot. Res., 2024

2024

[4] [4]

RT-1: Robot. Transformer for Real-World Control at Scale,

A. Brohan, N. Brown, and J. C. et al., “RT-1: Robot. Transformer for Real-World Control at Scale,” inProc. of Robot.: Sci. and Syst., Daegu, Republic of Korea, July 2023

2023

[5] [5]

Dasari, O

S. Dasari, O. Mees, S. Zhao, M. K. Srirama, and S. Levine, “The ingredients for robotic diffusion transformers,”arXiv preprint arXiv:2410.10088, 2024

work page arXiv 2024

[6] [6]

Film: visual reasoning with a general conditioning layer,

E. Perez, F. Strub, H. de Vries, V . Dumoulin, and A. Courville, “Film: visual reasoning with a general conditioning layer,” inProceedings of the Thirty-Second AAAI Conf. on Artificial Intelligence and Thirtieth Innovative Applications of Artificial Intelligence Conf. and Eighth AAAI Symposium on Educational Advances in Artificial Intelligence, ser. AAAI’...

2018

[7] [7]

Scil: Stage- Conditioned Imitation Learning for Multi-Stage Manipulation,

Z. Ouyang, K. Wang, J. Liu, H. Lu, and W. Zhang, “Scil: Stage- Conditioned Imitation Learning for Multi-Stage Manipulation,”IEEE Control Syst. Letters, vol. 9, pp. 913–918, 2025

2025

[8] [8]

An incremental hybrid impedance algorithm for stable force-position control of elastic materials,

Y . Zhang, D. Chen, W. He, A. E. P. Barcel´o, J. V . S. Luces, and Y . Hirata, “An incremental hybrid impedance algorithm for stable force-position control of elastic materials,”IEEE Trans. on Automat. Sci. and Eng., pp. 1–1, 2025

2025

[9] [9]

Goal-conditioned model simpli- fication for 1-d and 2-d deformable object manipulation,

S. Wang, M. Leonetti, and M. Dogar, “Goal-conditioned model simpli- fication for 1-d and 2-d deformable object manipulation,”IEEE Trans. on Robot., vol. 41, pp. 4023–4040, 2025

2025

[10] [10]

Sis: Seam-informed strategy for t-shirt unfolding,

X. Huang, A. Seino, F. Tokuda, A. Kobayashi, D. Chen, Y . Hirata, N. C. Tien, and K. Kosuge, “Sis: Seam-informed strategy for t-shirt unfolding,” IEEE Robot. and Automat. Letters, vol. 10, no. 7, pp. 7342–7349, 2025

2025

[11] [11]

RTFF: Random-to-Target Fabric Flattening Policy using Dual-Arm Manipulator

K. Tang, D. Bhattacharya, H. Xu, F. Tokuda, N. C. Tien, and K. Ko- suge, “Rtff: Random-to-target fabric flattening policy using dual-arm manipulator,” https://arxiv.org/abs/2510.00814, 2025

work page internal anchor Pith review Pith/arXiv arXiv 2025

[12] [12]

Graph- garment: Learning garment dynamics for bimanual cloth manipulation tasks,

W. Chen, K. Li, D. Lee, X. Chen, R. Zong, and P. Kormushev, “Graph- garment: Learning garment dynamics for bimanual cloth manipulation tasks,” https://arxiv.org/abs/2503.05817, 2025

work page arXiv 2025

[13] [13]

Bimanual deformable bag manipulation using a structure-of-interest based neural dynamics model,

P. Zhou, P. Zheng, J. Qi, C. Li, H.-Y . Lee, Y . Pan, C. Yang, D. Navarro- Alarcon, and J. Pan, “Bimanual deformable bag manipulation using a structure-of-interest based neural dynamics model,”IEEE/ASME Trans. on Mechatronics, vol. 30, no. 5, pp. 3254–3265, 2025

2025

[14] [14]

Deformpam: Data- Efficient Learning for Long-Horizon Deformable Object Manipulation Via Preference-Based Action Alignment,

W. Chen, H. Xue, F. Zhou, Y . Fang, and C. Lu, “Deformpam: Data- Efficient Learning for Long-Horizon Deformable Object Manipulation Via Preference-Based Action Alignment,” in2025 IEEE Int. Conf. on Robot. and Automat. (ICRA). IEEE, may 19 2025, pp. 6896–6903

2025

[15] [15]

Learning human-inspired force strategies for robotic assembly,

S. Scherzinger, A. Roennau, and R. Dillmann, “Learning human-inspired force strategies for robotic assembly,” in2023 IEEE 19th Int. Conf. on Automat. Sci. and Eng. (CASE), 2023, pp. 1–8

2023

[16] [16]

A system for imitation learning of contact-rich bimanual manipulation policies,

S. Stepputtis, M. Bandari, S. Schaal, and H. B. Amor, “A system for imitation learning of contact-rich bimanual manipulation policies,” in 2022 IEEE/RSJ Int. Conf. on Itell. Robots and Syst. (IROS), 2022, pp. 11 810–11 817

2022

[17] [17]

FILIC: Dual-Loop Force-Guided Imitation Learning with Impedance Torque Control for Contact-Rich Manipulation Tasks

H. Ge, Y . Jia, Z. Li, Y . Li, Z. Chen, R. Huang, and G. Zhou, “Filic: Dual- loop force-guided imitation learning with impedance torque control for contact-rich manipulation tasks,” https://arxiv.org/abs/2509.17053, 2025

work page internal anchor Pith review Pith/arXiv arXiv 2025

[18] [18]

A reduction of imitation learning and structured prediction to no-regret online learning,

S. Ross, G. Gordon, and D. Bagnell, “A reduction of imitation learning and structured prediction to no-regret online learning,” inProceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics, ser. Proceedings of Machine Learning Research, G. Gordon, D. Dunson, and M. Dud ´ık, Eds., vol. 15. Fort Lauderdale, FL, USA: PML...

2011

[19] [19]

Rac: Robot learning for long-horizon tasks by scaling recovery and correction,

Z. Hu, R. Wu, N. Enock, J. Li, R. Kadakia, Z. Erickson, and A. Kumar, “Rac: Robot learning for long-horizon tasks by scaling recovery and correction,” 2025

2025

[20] [20]

On the continuity of rotation representations in neural networks,

Y . Zhou, C. Barnes, J. Lu, J. Yang, and H. Li, “On the continuity of rotation representations in neural networks,” https://arxiv.org/abs/1812. 07035, 2020

2020

[21] [21]

Simplified offset and gravity compensation for wrist- mounted force/torque sensor with a garment manipulation application,

D. Chen, Y . Zhang, W. He, A. E. Petrilli Barcel ´o, J. V . Salazar Luces, and Y . Hirata, “Simplified offset and gravity compensation for wrist- mounted force/torque sensor with a garment manipulation application,” in2025 IEEE Int. Conf. on Robot. and Biomimetics (ROBIO), 2025

2025

[22] [22]

Forward dynamics compliance control (fdcc): A new approach to cartesian compliance for robotic manipulators,

S. Scherzinger, A. Roennau, and R. Dillmann, “Forward dynamics compliance control (fdcc): A new approach to cartesian compliance for robotic manipulators,” inIEEE/RSJ Int. Conf. on Intell. Robots and Syst. (IROS), 2017, pp. 4568–4575

2017

[23] [23]

Oculus reader: Robotic teleoperation interface,

F. E. Jedrzej Orbik, “Oculus reader: Robotic teleoperation interface,” https://github.com/rail-berkeley/oculus reader, 2021

2021

[24] [24]

Coordinated motion control of dual manipulators for handling a rigid object with non-negligible deformation,

K. Kosuge, K. Kamei, and T. Nammoto, “Coordinated motion control of dual manipulators for handling a rigid object with non-negligible deformation,” in2014 IEEE Int. Conf. on Robot. and Automat. (ICRA), 2014, pp. 5145–5151

2014

[25] [25]

Lerobot: State-of-the-art machine learning for real-world robotics in pytorch,

R. Cadene, S. Alibert, A. Soare, Q. Gallouedec, A. Zouitine, S. Palma, P. Kooijmans, M. Aractingi, M. Shukor, D. Aubakirova, M. Russi, F. Capuano, C. Pascal, J. Choghari, J. Moss, and T. Wolf, “Lerobot: State-of-the-art machine learning for real-world robotics in pytorch,” https://github.com/huggingface/lerobot, 2024. 9 Dayuan Chen(Graduate Student Member...

2024