Research map / Reinforcement learning / Visual and representation RL
AcoustoReinforce: Multi-Particle Acoustophoretic Path Planning with Deep Reinforcement Learning
Pengyuan Wei, Giorgos Christopoulos, Zhouyang Shen, Jincheng Wang, Joshua Mukherjee, Ryuji Hirayama, Sriram Subramanian, Prateek Mittal
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use AcoustoReinforce instead of S2M2 | multi-particle path planning · Simulation, 10 particles | success rate: 0.8920 -> 0.9830 | Improved |
| use AcoustoReinforce instead of CBS | multi-particle path planning · Simulation, 10 particles | success rate: 0.9370 -> 0.9830 | Improved |
| use AcoustoReinforce instead of S2M2 and CBS at low particle counts | multi-particle path planning · Real levitator, 4 particles, Vmax 0.05 to 0.15 m/s | success rate: 1.000 for all three | No clear effect |
| use AcoustoReinforce instead of S2M2 | multi-particle path planning · Real levitator, 8 particles, Vmax 0.20 m/s | success rate: 0.2000 -> 0.4600 | Improved |
| use AcoustoReinforce instead of CBS | multi-particle path planning · Real levitator, 10 particles, Vmax 0.20 m/s | success rate: 0.0000 for all methods | No clear effect |
| use AcoustoReinforce instead of CBS | multi-particle path planning · Real levitator, 6 particles, Vmax 0.10 m/s | success rate: 0.9000 -> 0.9400 | Improved |
| use AcoustoReinforce instead of CBS | multi-particle path planning · Real levitator, 6 particles, Vmax 0.05 m/s | success rate: 1.000 -> 0.9900 | Hurt |
| use AcoustoReinforce instead of S2M2 and CBS | multi-particle path planning · Gor'kov potential at turning points | potential (lower is better): reductions of up to 20% | Improved |
| add turning-point Gor'kov optimization to the policy (ablation) | AcoustoReinforce · Gor'kov potential at turning points | potential (lower is better): policy alone improves by approximately 5% | Improved |
| increase particle count from 4 to 10 on difference between planners in Gor'kov potential | AcoustoReinforce · Gor'kov potential at turning points | potential: differences decrease with more particles | Mixed |
| increase Vmax from 0.05 to 0.20 m/s | AcoustoReinforce · Real levitator, 8 particles | success rate: 0.9900 -> 0.4600 | Hurt |
Closest papers by meaning
- Learning Particle Dynamics for Manipulating Rigid Bodies, Deformable Objects, and Fluids (ICLR 2019 · similarity 0.82)
- DiffPhyCon: A Generative Approach to Control Complex Physical Systems (NeurIPS 2024 · similarity 0.81)
- Move2Hear: Active Audio-Visual Source Separation (ICCV 2021 · similarity 0.81)
- Stabilizing Reinforcement Learning in Differentiable Multiphysics Simulation (ICLR 2025 · similarity 0.81)
- NeuMA: Neural Material Adaptor for Visual Grounding of Intrinsic Dynamics (NeurIPS 2024 · similarity 0.80)
- Rapidly Adapting Policies to the Real-World via Simulation-Guided Fine-Tuning (ICLR 2025 · similarity 0.80)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
