Self-supervised learning of the biologically-inspired obstacle avoidance of hexapod walking robot
Source abstract
Abstract
In this paper, we propose an integrated biologically inspired visual collision avoidance approach that is deployed on a real hexapod walking robot. The proposed approach is based on the Lobula giant movement detector (LGMD), a neural network for looming stimuli detection that can be found in visual pathways of insects, such as locusts. Although a superior performance of the LGMD in the detection of intercepting objects has been shown in many collision avoiding scenarios, its direct integration with motion control is an unexplored topic. In our work, we propose to utilize the LGMD neural network for visual interception detection with a central pattern generator (CPG) for locomotion control of a hexapod walking robot that are combined in the controller based on the long short-term memory (LSTM) recurrent neural network. Moreover, we propose self-supervised learning of the integrated controller to autonomously find a suitable setting of the system using a realistic robotic simulator. Thus, individual neural networks are trained in a simulation to enhance the performance of the controller that is then experimentally verified with a real hexapod walking robot in both collision and interception avoidance scenario and navigation in a cluttered environment.
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: Robotic Locomotion and Control · Zebrafish Biomedical Research Applications · Biomimetic flight and propulsion mechanisms
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (38)
Show source-provided references
1. 10.1109/ROBOT.2005.157057910.1109/ROBOT.2005.1570579 ↗
2. 10.1016/S0921-8890(99)00063-910.1016/S0921-8890(99)00063-9 ↗
3. Braitenberg V · Vehicles: Experiments in Synthetic Psychology · 1986
4. Chalupka K · Bioinspir. Biomim. · 2016 · 1110.1088/1748-3190/11/4/046008 ↗
5. Chollet F · 2015
6. Čížek P · CEUR Workshop Proc. Information Technologies—Application and Theory · 2017 · 1885 · 65
7. 10.1109/IJCNN.2017.796591410.1109/IJCNN.2017.7965914 ↗
8. 10.1016/j.cub.2012.01.00710.1016/j.cub.2012.01.007 ↗
9. 10.1152/jn.00073.200910.1152/jn.00073.2009 ↗
10. 10.1007/s00422-006-0129-x10.1007/s00422-006-0129-x ↗
11. 10.5244/C.30.610.5244/C.30.6 ↗
12. 10.1109/IROS.2017.820624710.1109/IROS.2017.8206247 ↗
13. Hartbauer M · Bioinspir. Biomim. · 2017 · 1210.1088/1748-3190/aa5993 ↗
14. 10.1162/neco.1997.9.8.173510.1162/neco.1997.9.8.1735 ↗
15. 10.1017/S026357471400028910.1017/S0263574714000289 ↗
16. 10.1016/j.neunet.2008.03.01410.1016/j.neunet.2008.03.014 ↗
17. Judge S · J. Exp. Biol. · 1997 · 200 · 220910.1242/jeb.200.16.2209 ↗
18. Kelchtermans K · CoRR · 2017
19. 10.1109/LRA.2018.279564310.1109/LRA.2018.2795643 ↗
20. 10.1007/BF0031951410.1007/BF00319514 ↗
21. Miller B D · Bioinspir. Biomim. · 2015 · 10
22. 10.1109/RoMoCo.2015.721974210.1109/RoMoCo.2015.7219742 ↗
23. Muller U · Advances in Neural Information Proc. Systems · 2006 · 739
24. 10.1007/978-3-030-14984-0_1710.1007/978-3-030-14984-0_17 ↗
25. 10.1109/ICRA.2011.597956110.1109/ICRA.2011.5979561 ↗
26. 10.1523/JNEUROSCI.5259-05.200610.1523/JNEUROSCI.5259-05.2006 ↗
27. 10.1109/IROS.2013.669652010.1109/IROS.2013.6696520 ↗
28. Salt L · CoRR · 2017
29. Sutskever I · 2013
30. Szadkowski R J · CEUR Workshop Proc. Information Technologies—Application and Theory · 2018 · 2203 · 116
31. Wilson D M · J. Exp. Biol. · 1961 · 38 · 47110.1242/jeb.38.2.471 ↗
32. 10.1109/TCYB.2015.247923710.1109/TCYB.2015.2479237 ↗
33. 10.1109/TIE.2016.256948910.1109/TIE.2016.2569489 ↗
34. 10.1109/TNN.2006.87328610.1109/TNN.2006.873286 ↗
35. 10.1109/TAMD.2013.225505010.1109/TAMD.2013.2255050 ↗
36. 10.1016/j.cub.2012.05.02410.1016/j.cub.2012.05.024 ↗
37. 10.1109/IJCNN.2018.848929810.1109/IJCNN.2018.8489298 ↗
38. 10.1109/TCST.2017.269272710.1109/TCST.2017.2692727 ↗