CrowdGAIL: A spatiotemporal aware method for agent navigation
Agent navigation has been a crucial task in today's service and automated factories. Many efforts are to set specific rules for agents in a certain scenario to regulate the agent's behaviors. However, not all situations could be in advance considered, which might lead to terrible performan...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
AIMS Press
2023-01-01
|
| Series: | Electronic Research Archive |
| Subjects: | |
| Online Access: | https://www.aimspress.com/article/doi/10.3934/era.2023057?viewType=HTML |
| _version_ | 1827954284802605056 |
|---|---|
| author | Longchao Da Hua Wei |
| author_facet | Longchao Da Hua Wei |
| author_sort | Longchao Da |
| collection | DOAJ |
| description | Agent navigation has been a crucial task in today's service and automated factories. Many efforts are to set specific rules for agents in a certain scenario to regulate the agent's behaviors. However, not all situations could be in advance considered, which might lead to terrible performance in a real-world application. In this paper, we propose CrowdGAIL, a method to learn from expert behaviors as an instructing policy, can train most 'human-like' agents in navigation problems without manually setting any reward function or beforehand regulations. First, the proposed model structure is based on generative adversarial imitation learning (GAIL), which imitates how humans take actions and move toward the target to a maximum extent, and by comparison, we prove the advantage of proximal policy optimization (PPO) to trust region policy optimization, thus, GAIL-PPO is what we base. Second, we design a special Sequential DemoBuffer compatible with the inner long short-term memory structure to apply spatiotemporal instruction on the agent's next step. Third, the paper demonstrates the potential of the model with an integrated social manner in a multi-agent scenario by considering human collision avoidance as well as social comfort distance. At last, experiments on the generated dataset from CrowdNav verify how close our model would act like a human being in the trajectory aspect and also how it could guide the multi-agents by avoiding any collision. Under the same evaluation metrics, CrowdGAIL shows better results compared with classic Social-GAN. |
| first_indexed | 2024-04-09T14:27:45Z |
| format | Article |
| id | doaj.art-f365b90091bd4be19a66ba2eadd670f9 |
| institution | Directory Open Access Journal |
| issn | 2688-1594 |
| language | English |
| last_indexed | 2024-04-09T14:27:45Z |
| publishDate | 2023-01-01 |
| publisher | AIMS Press |
| record_format | Article |
| series | Electronic Research Archive |
| spelling | doaj.art-f365b90091bd4be19a66ba2eadd670f92023-05-04T01:21:11ZengAIMS PressElectronic Research Archive2688-15942023-01-013121134114610.3934/era.2023057CrowdGAIL: A spatiotemporal aware method for agent navigationLongchao Da0Hua Wei1Department of Informatics, New Jersey Institute of Technology, Newark, New Jersey 07102, USADepartment of Informatics, New Jersey Institute of Technology, Newark, New Jersey 07102, USAAgent navigation has been a crucial task in today's service and automated factories. Many efforts are to set specific rules for agents in a certain scenario to regulate the agent's behaviors. However, not all situations could be in advance considered, which might lead to terrible performance in a real-world application. In this paper, we propose CrowdGAIL, a method to learn from expert behaviors as an instructing policy, can train most 'human-like' agents in navigation problems without manually setting any reward function or beforehand regulations. First, the proposed model structure is based on generative adversarial imitation learning (GAIL), which imitates how humans take actions and move toward the target to a maximum extent, and by comparison, we prove the advantage of proximal policy optimization (PPO) to trust region policy optimization, thus, GAIL-PPO is what we base. Second, we design a special Sequential DemoBuffer compatible with the inner long short-term memory structure to apply spatiotemporal instruction on the agent's next step. Third, the paper demonstrates the potential of the model with an integrated social manner in a multi-agent scenario by considering human collision avoidance as well as social comfort distance. At last, experiments on the generated dataset from CrowdNav verify how close our model would act like a human being in the trajectory aspect and also how it could guide the multi-agents by avoiding any collision. Under the same evaluation metrics, CrowdGAIL shows better results compared with classic Social-GAN.https://www.aimspress.com/article/doi/10.3934/era.2023057?viewType=HTMLgailsocial awarenessagent navigation |
| spellingShingle | Longchao Da Hua Wei CrowdGAIL: A spatiotemporal aware method for agent navigation Electronic Research Archive gail social awareness agent navigation |
| title | CrowdGAIL: A spatiotemporal aware method for agent navigation |
| title_full | CrowdGAIL: A spatiotemporal aware method for agent navigation |
| title_fullStr | CrowdGAIL: A spatiotemporal aware method for agent navigation |
| title_full_unstemmed | CrowdGAIL: A spatiotemporal aware method for agent navigation |
| title_short | CrowdGAIL: A spatiotemporal aware method for agent navigation |
| title_sort | crowdgail a spatiotemporal aware method for agent navigation |
| topic | gail social awareness agent navigation |
| url | https://www.aimspress.com/article/doi/10.3934/era.2023057?viewType=HTML |
| work_keys_str_mv | AT longchaoda crowdgailaspatiotemporalawaremethodforagentnavigation AT huawei crowdgailaspatiotemporalawaremethodforagentnavigation |