作者: Grzegorz Cielniak , Tom Duckett , Feras Dayoub
DOI:
关键词: Robot 、 Computer science 、 Mobile service 、 Mobile robot navigation 、 Artificial intelligence 、 Mobile robot 、 Human–computer interaction 、 Memory model 、 Modal 、 Semantic memory 、 Term (time)
摘要: Throughout a lifetime of operation, mobile service robot needs to acquire, store and update its knowledge working environment. This includes the ability identify track objects in different places, as well using this information for interaction with humans. paper introduces long-term updating mechanism, inspired by modal model human memory, enable maintain changing The memory is integrated hybrid map that represents global topology local geometry environment, respective 3D location objects. We aim use help humans suggesting most likely locations specific map. An experiment omni-directional vision demonstrates movements several dynamic environment over an extended period time.