This thesis presents a perceptual system for a humanoid robot that integrates abilities such as object localization and recognition with the deeper developmental machinery required to forge those competences out of raw physical experiences. It shows that a robotic platform can build up and maintain a system for object localization, segmentation, and recognition, starting from very little. What the robot starts with is a direct solution to achieving figure/ground separation: it simply `pokes around' in a region of visual ambiguity and watches what happens. If the arm passes through an area, that area is recognized as free space. If the arm collides with an object, causing it to move, the robot can use that motion to segment the object from the background. Once the robot can acquire reliable segmented views of objects, it learns from them, and from then on recognizes and segments those objects without further contact. Both low-level and high-level visual features can also be learned in this way, and examples are presented for both: orientation detection and affordance recognition, respectively.

Abstract — To perform successful manipulation, robots depend on information about objects in their environment. In unstructured environments, such information cannot be given to the robot a priori. It is thus critical for the robot to be able to continuously acquire task-specific information about objects. Towards this goal, we present a robust perceptual skill for identifying, tracking, and segmenting objects in a cluttered environment. We increase the robot’s perceptual capabilities by closely coupling them with the robot’s manipulation skills. The robot’s interaction with objects in the environment creates a perceptual signal, i.e. motion, that renders segmentation and tracking robust and reliable. In addition, the resulting perceptual signal reveals the type of segmentation most relevant to manipulation, namely a segmentation of rigidly connected physical bodies. We demonstrate our approach with experiments on a real world mobile manipulation platform with multiple objects in a cluttered scene. I.

Humans use a set of exploratory procedures to examine object properties through grasping and touch. Our goal is to exploit similar methods with a humanoid robot to enable developmental learning about manipulation. We use a compliant robot hand to find objects without prior knowledge of their presence or location, and then tap those objects with a finger. This behavior lets the robot generate and collect samples of the contact sound produced by impact with that object. We demonstrate the feasibility of recognizing objects by their sound, and relate this to human performance under situations analogous to that of the robot. 1.

For a robot to be capable of development, it must be able to explore its environment and learn from its experiences. It must find (or create) opportunities to experience the unfamiliar in ways that reveal properties valid beyond the immediate context. In this paper, we develop a novel method for using the rhythm of everyday actions as a basis for identifying the characteristic appearance and sounds associated with objects, people, and the robot itself. Our approach is to identify and segment groups of signals in individual modalities (sight, hearing, and proprioception) based on their rhythmic variation, then to identify and bind causally-related groups of signals across different modalities. By including proprioception as a modality, this cross-modal binding method applies to the robot itself, and we report a series of experiments in which the robot learns about the characteristics of its own body. 1.

For a robot to be capable of development, it must be able to explore its environment and learn from its experiences. It must find (or create) opportunities to experience the unfamiliar in ways that reveal properties valid beyond the immediate context. In this paper, we develop a novel method for using the rhythm of everyday actions as a basis for identifying the characteristic appearance and sounds associated with objects, people, and the robot itself. Our approach is to identify and segment groups of signals in individual modalities (sight, hearing, and proprioception) based on their rhythmic variation, then to identify and bind causally-related groups of signals across different modalities. By including proprioception as a modality, this cross-modal binding method applies to the robot itself, and we report a series of experiments in which the robot learns about the characteristics of its own body. 1

Abstract. This paper presents a new embodied approach for object segmentation by a humanoid robot. It relies on interactions with a human teacher that drives the robot through the process of segmenting objects from arbitrarily complex, non-static images. Objects from a large spectrum of different scenarios were successfully segmented by the proposed algorithms. 1

An integrated system for teaching new visually grounded words to a

For a robot to be capable of development, it must be able to explore its environment and learn from its experiences. It must find (or create) opportunities to experience the unfamiliar in ways that reveal properties valid beyond the immediate context. In this paper, we develop a novel method for using the rhythm of everyday actions as a basis for identifying the characteristic appearance and sounds associated with objects, people, and the robot itself. Our approach is to identify and segment groups of signals in individual modalities (sight, hearing, and proprioception) based on their rhythmic variation, then to identify and bind causally-related groups of signals across different modalities. By including proprioception as a modality, this cross-modal binding method applies to the robot itself, and we report a series of experiments in which the robot learns about the characteristics of its own body. 1.