Asil Kaan Bozcuoglu

Learning Adjectives and Nouns from Affordances on the iCub Humanoid Robot

This article studies how a robot can learn nouns and adjectives in language. Towards this end, we extended a framework that enabled robots to learn affordances from its sensorimotor interactions, to learn nouns and adjectives using labeling from humans. Specifically, an iCub humanoid robot interacted with a set of objects (each labeled with a set of adjectives and a noun) and learned to predict the effects (as labeled with a set of verbs) it can generate on them with its behaviors. Different from appearance-based studies that directly link the appearances of objects to nouns and adjectives, we first predict the affordances of an object through a set of Support Vector Machine classifiers which provided a functional view of the object. Then, we learned the mapping between these predicted affordance values and nouns and adjectives. We evaluated and compared a number of different approaches towards the learning of nouns and adjectives on a small set of novel objects.

Open robotics research using web-based knowledge services

In this paper we discuss how the combination of modern technologies in “big data” storage and management, knowledge representation and processing, cloud-based computation, and web technology can help the robotics community to establish and strengthen an open research discipline. We describe how we made the demonstrator of a EU project review openly available to the research community. Specifically, we recorded episodic memories with rich semantic annotations during a pizza preparation experiment in autonomous robot manipulation. Afterwards, we released them as an open knowledge base using the cloud- and web-based robot knowledge service OPENEASE. We discuss several ways on how this open data can be used to validate our experimental reports and to tackle novel challenging research problems.

The Exchange of Knowledge Using Cloud Robotics

To enable robots to perform human-level tasks flexibly in varying conditions, we need a mechanism that allows them to exchange knowledge between themselves for crowd-sourcing the knowledge gap problem. One approach to achieve this is to equip a cloud application with a range of encyclopedic knowledge (i.e. ontologies) and execution logs of different robots performing the same tasks in different environments. In this paper, we show how knowledge exchange between robots can be done using OPENEASE as the cloud application. We equipped OPENEASE with ontologies about the kitchen domain, execution logs of three robots operating in two different kitchens, and semantic descriptions of both environments. By addressing two different use cases, we show that two PR2 robots and one Fetch robot can successfully adapt each others plan parameters and sub symbolic data to the experiments that they are conducting.

Heterogeneous Ontologies and Hybrid Reasoning for Service Robotics: The EASE Framework

As robots are expected to accomplish human-level manipulation tasks, the demand for formal knowledge representation techniques and reasoning for robots increases dramatically. In this paper we describe how to make use of heterogeneous ontologies in service robotics. To illustrate the vision, we take the action of pouring as an example.

Rendering semantically-annotated experiment videos out of robot memories

Towards life-long learning schemes for cognitive robots, having a human-like episodic memory structure and management is an important capability. By having this, they will have data from past experiences to carry such a learning process. In this paper, we show how this kind of detailed robot memories, such as the one described in [1], can be used to generate a video of the episode with semantic annotations. This methodology does not only prove that we have an adequately-detailed episodic memory structure for robots but also becomes a comprehensive tool for roboticists while analyzing, diagnosing and debugging how autonomous robots have behaved under certain conditions.

Traversability on a simple humanoid: What did I just trip over?

The notion of affordance has taken the attention of roboticists in recent years. Previously we had used this concept to learn and perceive the traversability of a mobile robot platform. In this paper, we have shown how a simplistic humanoid robot equipped with time-of-flight ultrasonic sensor can learn traversability affordance. In addition to this, we have demonstrated it can infer how sensory data history affect this affordance by merging previously sensed data with the current data via a sliding data window concatenating recent history of sensor activity. This sliding window approach improves the performance of the system in the cases where the object is invisible at the time of collision. We performed several experiments in which the robot attained the ability to generalize what it has already learned by performing the move forward behavior robustly in cluttered environments with novel objects albeit noisy range measurements.