Collective motion model inspired by fish school based on deep attention mechanism
Source abstract
Abstract
Abstract Collective intelligence in biological groups can be employed to inspire the control of artificial complex systems, such as swarm robotics. However, modeling for the social interactions between individuals is still a challenging task. Without loss of generality, we propose a deep attention network model that incorporates the principles of biological Hard Attention mechanisms, that means an individual only pay attention to one or two neighbors for collective motion decision in large group. The model is trained by the collective movement data of five rummy-nose tetra fish (Hemigrammus rhodostomus). The structure of the model enforces individual agents to consider information from at most two neighboring agents. Meanwhile, the model can reveal hidden locations, where highly influential neighbors frequently appear. These findings demonstrate that the proposed Hard Attention Model aligns with the information processing mechanisms, which is observed in fish schooling. Experimental results indicate that the model exhibits a strong ability to decouple sparse information for collective movement with robust metrics. It can also perform excellent scalability in different group sizes. The simulation and real robots experiment show that the model provides a powerful tool for analyzing multi-level behaviors in complex systems and offers significant insights for the distributed control of swarm robotics.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Water Quality Monitoring Technologies
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (56)
Show source-provided references
1. Application of 3D animation cluster system based on artificial intelligence and machine learning10.1155/2022/2904607 ↗
2. Collective motion of cells modeled as ring polymers10.1039/d1sm01640g ↗
3. Testing a thermodynamic approach to collective animal behavior in laboratory fish schools10.1103/physreve.101.062605 ↗
4. Energy management of echelon flying northern bald ibises with different wingspans and variable wingtip spacing10.1007/s42235-021-00107-7 ↗
5. Animal migration to northern latitudes: environmental changes and increasing threats10.1016/j.tree.2021.08.010 ↗
6. Self-propelled predator-prey of swarming Janus micromotors10.1016/j.isci.2023.106112 ↗
7. Contagion dynamics in self-organized systems of self-propelled agents10.1038/s41598-022-06083-0 ↗
8. A review on collective behavior modeling and simulation: building a link between cognitive psychology and physical action10.1007/s10489-023-04924-7 ↗
9. The behavioral model and game theory10.1057/s41599-019-0265-2 ↗
10. Collective dynamics of active Brownian particles in three spatial dimensions: a predictive field theory10.1103/physrevresearch.2.033241 ↗
11. Swarm and swarm intelligence—introductory study into collective behaviour of natural and artificial systems10.58532/v3bgnc2p9ch2 ↗
12. Collective neural network behavior in a dynamically driven disordered system of superconducting loops10.1073/pnas.2314995121 ↗
13. Collective motion10.1016/j.physrep.2012.03.004 ↗
14. The simulation of the movement of fish schools10.1016/s0022-5193(05)80681-2 ↗
15. Embodied neuromorphic intelligence10.1038/s41467-022-28487-2 ↗
16. Embodied intelligence via learning and evolution10.1038/s41467-021-25874-z ↗
17. Visual ecology of fish–a review with special reference to percids10.1016/j.semcdb.2020.05.027 ↗
18. Computational and robotic modeling reveal parsimonious combinations of interactions between individuals in schooling fish10.1371/journal.pcbi.1007194 ↗
19. Voluntary initiation of movement: multifunctional integration of subjective agency10.3389/fpsyg.2015.00688 ↗
20. Brain dynamics altered by photic stimulation in patients with Alzheimer’s disease and mild cognitive impairment10.3390/e23040427 ↗
21. Novel type of phase transition in a system of self-driven particles10.1103/physrevlett.75.1226 ↗
22. The physics of the Vicsek model10.1140/epjst/e2016-60066-8 ↗
23. Cooperative event-triggered fuzzy-neural multivariable control with multitask learning for municipal solid waste incineration process10.1109/tii.2023.3264108 ↗
24. An improved Vicsek model of swarm based on remote neighbors strategy10.1016/j.physa.2021.126553 ↗
25. Motional consensus of self-propelled particles10.1038/s41598-023-35238-w ↗
26. Spatial models of bistability in biological collectives10.1109/cdc.2007.4434523 ↗
27. Interaction ruling animal collective behavior depends on topological rather than metric distance: evidence from a field study10.1073/pnas.0711437105 ↗
28. Convergence rate for discrete-time multiagent systems with time-varying delays and general coupling coefficients10.1109/tnnls.2015.2473690 ↗
29. Network anomaly detection using a graph neural network10.32920/26883595 ↗
30. GPENs: graph data learning with graph propagation-embedding networks10.1109/tnnls.2021.3120100 ↗
31. Extracting symbolic models of collective behaviors with graph neural networks and macro-micro evolution10.1007/978-3-031-20176-9_12 ↗
32. The impact of individual perceptual and cognitive factors on collective states in a data-driven fish school model10.1371/journal.pcbi.1009437 ↗
33. Dynamical order and many-body correlations in zebrafish show that three is a crowd10.1038/s41467-024-46426-1 ↗
34. Edge-cloud solutions for big data analysis and distributed machine learning—110.1016/j.future.2024.05.023 ↗
35. Deep learning on medical image analysis10.1049/cit2.12356 ↗
36. Big data user behaviour prediction model incorporating deep learning10.2478/amns-2024-1450 ↗
37. Collective density clustering for coherent motion detection10.1109/TMM.2017.2771477 ↗
38. Graph SLAM built over point clouds matching for robot localization in tunnels10.3390/s21165340 ↗
39. From shadows to light: a swarm robotics approach with onboard control for seeking dynamic sources in constrained environments10.1109/lra.2023.3331897 ↗
40. Social networks and models for collective motion in animals10.1007/s00265-010-1111-0 ↗
41. Advancements and challenges of information integration in swarm robotics10.1109/cis-ram55796.2023.10370011 ↗
42. Intelligent control of bionic collective motion based on deep learning10.13195/j.kzyjc.2020.0071 ↗
43. Transfer control of swarm robotics motion reinforcement employing fish schooling emergency mechanism10.13195/j.kzyjc.2021.1554 ↗
44. Collective memory and spatial sorting in animal groups10.1006/jtbi.2002.3065 ↗
45. Deep attention networks reveal the rules of collective motion in zebrafish10.1371/journal.pcbi.1007354 ↗
46. Analysis model for revealing mechanism of internal information coupling in biological collective systems based on deep network10.13195/j.kzyjc.2022.1079 ↗
47. A data-driven method for reconstructing and modelling social interactions in moving animal groups10.1098/rstb.2019.0380 ↗
48. Draft genome of the mirrorwing flyingfish (Hirundichthys speculiger)10.3389/fgene.2021.695700 ↗
49. Disentangling and modeling interactions in fish with burst-and-coast swimming reveal distinct alignment and attraction behaviors10.1371/journal.pcbi.1005933 ↗
50. Attention is all you need. Advances in neural information processing systems
51. Transvpr: transformer-based place recognition with multi-level attention aggregation10.1109/cvpr52688.2022.01328 ↗
52. Vit-ret: vision and recurrent transformer neural networks for human activity recognition in videos10.1109/access.2023.3293813 ↗
53. Quantification of the influence of drugs on zebrafish larvae swimming kinematics and energetics10.7717/peerj.8374 ↗
54. Acoustic and visual stimuli combined promote stronger responses to aerial predation in fish10.1093/beheco/arab043 ↗
55. Bio-inspired dynamic collective choice in large-population systems: a robust mean-field game perspective10.1109/tnnls.2020.3027428 ↗
56. Magnetically programmable cuboids for 2D locomotion and collaborative assembly10.1109/iros45743.2020.9341022 ↗