Learning dynamic cognitive map with autonomous navigation

被引：0
作者：
de Tinguy, Daria [1 ]
Verbelen, Tim [2 ]
Dhoedt, Bart [1 ]
机构：
[1] Department of Engineering and Architecture, Ghent University, IMEC, Ghent
[2] Verses, Los Angeles, CA
来源：
Frontiers in Computational Neuroscience | 2024年 / 18卷
关键词：
active inference; autonomous navigation; cognitive map; dynamic mapping; knowledge learning; structure learning;
D O I：
10.3389/fncom.2024.1498160
中图分类号：
学科分类号：
摘要：
Inspired by animal navigation strategies, we introduce a novel computational model to navigate and map a space rooted in biologically inspired principles. Animals exhibit extraordinary navigation prowess, harnessing memory, imagination, and strategic decision-making to traverse complex and aliased environments adeptly. Our model aims to replicate these capabilities by incorporating a dynamically expanding cognitive map over predicted poses within an active inference framework, enhancing our agent's generative model plasticity to novelty and environmental changes. Through structure learning and active inference navigation, our model demonstrates efficient exploration and exploitation, dynamically expanding its model capacity in response to anticipated novel un-visited locations and updating the map given new evidence contradicting previous beliefs. Comparative analyses in mini-grid environments with the clone-structured cognitive graph model (CSCG), which shares similar objectives, highlight our model's ability to rapidly learn environmental structures within a single episode, with minimal navigation overlap. Our model achieves this without prior knowledge of observation and world dimensions, underscoring its robustness and efficacy in navigating intricate environments. Copyright © 2024 de Tinguy, Verbelen and Dhoedt.
引用
收藏
相关论文
共 59 条
  • [11] de Maele T.V., Dhoedt B., Verbelen T., Pezzulo G., “Integrating cognitive map learning and active inference for planning in ambiguous environments,”, International Workshop on Active Inference, pp. 204-217, (2023)
  • [12] de Tinguy D., Van de Maele T., Verbelen T., Dhoedt B., Spatial and temporal hierarchy for autonomous navigation using active inference in minigrid environment, Entropy, 26, (2024)
  • [13] Dedieu A., Lehrach W., Zhou G., George D., Lzaro-Gredilla M., Learning cognitive maps from transformer representations for efficient planning in partially observed environments, arXiv preprint arXiv:2401.05946, (2024)
  • [14] Edvardsen V., Bicanski A., Burgess N., Navigating with grid and place cells in cluttered environments, Hippocampus, 30, pp. 220-232, (2019)
  • [15] Eichenbaum H., The hippocampus as a cognitive map – of social space, Neuron, 87, pp. 9-11, (2015)
  • [16] Epstein R., Patai E.Z., Julian J., Spiers H., The cognitive map in humans: spatial navigation and beyond, Nat. Neurosci, 20, pp. 1504-1513, (2017)
  • [17] Foo P., Warren W., Duchon A., Tarr M., Do humans integrate routes into a cognitive map? Map- versus landmark-based navigation of novel shortcuts. Journal of experimental psychology, Learn. Memory Cogn, 31, pp. 195-215, (2005)
  • [18] Friston K., Life as we know it, J. R. Soc. Interface, 10, (2013)
  • [19] Friston K., FitzGerald T., Rigoli F., Schwartenbeck P., Doherty J.O., Pezzulo G., Active inference and learning, Neurosci. Biobehav. Rev, 68, pp. 862-879, (2016)
  • [20] Friston K., Moran R.J., Nagai Y., Taniguchi T., Gomi H., Tenenbaum J., World model learning and inference, Neural Netw, 144, pp. 573-590, (2021)
123456