Expectation-driven autonomous learning and interaction system

Abstract

Abstract-We introduce our latest autonomous learning and interaction system instance ALIS 2. It comprises different sensing modalities for visual (depth blobs, planar surfaces, motion) and auditory (speech, localization) signals and selfcollision free behavior generation on the robot ASIMO. The system design emphasizes the split into a completely autonomous reactive layer and an expectation generation layer. Different feature channels can be classified and named with arbitrary speech labels in on-line learning sessions. The feasibility of the proposed approach is shown by interaction experiments.

A computer vision integration model for a multi-modal cognitive system

“…The visuallinguistic instantiation we show in this work is just one example of possible cross-modal combinations. In this sense a similar approach is followed in [3]. Focusing to visuallinguistic integration, we see the main advantage of our work in more diversified, autonomous and responsive crossmodal learning mechanisms (implicit and explicit learning, clarification).…”

Self-Understanding and Self-Extension: A Systems and Representational Approach

“…Different systems focus on different aspects of the problem, such as the system architecture and integration [68], [69], [71], learning [66], [67], [71], or social interaction [70]. Our work focuses on the integration of visual and linguistic information by forming beliefs about the state of the world; these beliefs are used in the continuous learning process for updating the current representations.…”

A system for interactive learning in dialogue with a tutor

“…Different systems focus on different aspects of this problem, such as the system architecture and integration [3], [4], [6], learning [1], [2], [6], [7], or social interaction [5]. Our work focuses on the integration of visual perception and processing of linguistic information by forming beliefs about the state of the world; these beliefs are then used in the learning process for updating the current representations.…”

“…Interactive continuous learning using information obtained from vision and language is a desirable property of any cognitive system, therefore several systems have been developed that address this issue (e.g., [1], [2], [3], [4], [5], [6], [7]). Different systems focus on different aspects of this problem, such as the system architecture and integration [3], [4], [6], learning [1], [2], [6], [7], or social interaction [5].…”