Deep RL guides fish schools with virtual agents

Researchers have developed a deep reinforcement learning framework to guide schools of fish using virtual agents. The system employs Proximal Policy Optimization (PPO) to train policies that can be deployed in real-world experiments, interacting with live fish. A composite reward function was designed to balance directional guidance with cohesion, and experiments showed that a white background and larger stimulus sizes were most effective for guidance. The study also found that guidance efficacy decreased with larger group sizes and that multiple agents did not improve results. AI

IMPACT Demonstrates novel applications of reinforcement learning in biological systems, potentially influencing future bio-robotics research.

RANK_REASON The cluster contains an academic paper detailing a novel application of deep reinforcement learning. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• alphaXiv
• arXiv
• CatalyzeX
• DagsHub
• Gotit.pub
• Hiroaki Kawashima
• Hugging Face
• Petitella bleheri
• Proximal Policy Optimization
• ScienceCast

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →