SASRA: Semantically-aware Spatio-temporal Reasoning Agent for Vision-and-Language Navigation in Continuous Environments
2022 26th International Conference on Pattern Recognition (ICPR)（2022）
Key words
align map,autonomous agent,classical semantic mapping techniques,continuous 3D nvironments,cross-modal grounding,end-to-end learning-based VLN methods,hybrid transformer-recurrence model,language modalities,learning-based method,long-horizon simulation environment show,natural language instructions,raw visual observations,semantic spatio-temporal reasoning capabilities,semantically-aware spatio-temporal reasoning agent,temporal semantic memory,top-down local ego-centric semantic map,vision-and-language navigation,VLN policy
AI Read Science
Must-Reading Tree
Example
Generate MRT to find the research sequence of this paper
Chat Paper
Summary is being generated by the instructions you defined