Abstract
Long-term human-robot interaction requires the continuous acquisition of knowledge. This ability is referred to as lifelong learning (LL). LL is a long-standing challenge in machine learning due to catastrophic forgetting, which states that continuously learning from novel experiences leads to a decrease in the performance of previously acquired knowledge. Two recently published LL approaches are the Growing Dual-Memory (GDM) and the Self-organizing Incremental Neural Network+ (SOINN+). Both are growing neural networks that create new neurons in response to novel sensory experiences. The latter approach shows state-of-the-art clustering performance on sequentially available data with low memory requirements regarding the number of nodes. However, classification capabilities are not investigated. Two novel contributions are made in our research paper: (I) An extended SOINN+ approach, called associative SOINN+ (A-SOINN+), is proposed. It adopts two main properties of the GDM model to facilitate classification. (II) A new LL object recognition dataset (v-NICO-World-LL) is presented. It is recorded in a nearly photorealistic virtual environment, where a virtual humanoid robot manipulates 100 different objects belonging to 10 classes. Real-world and artificially created background images, grouped into four different complexity levels, are utilized. The A-SOINN+ reaches similar state-of-the-art classification accuracy results as the best GDM architecture of this work and consists of 30 to 350 times fewer neurons, evaluated on two LL object recognition datasets, the novel v-NICO-World-LL and the well-known CORe50. Furthermore, we observe an approximately 268 times lower training time. These reduced numbers result in lower memory and computational requirements, indicating higher suitability for autonomous social robots with low computational resources to facilitate a more efficient LL during long-term human-robot interactions.
Highlights
Social robots that interact with humans in their everyday lives are exposed to a dynamic and challenging environment
We focus on analyzing this constraint as it is directly linked to the two criteria, which we apply to evaluate the different approaches: the task performance and the number of created nodes
Both A-SOINN+ and Growing Dual-Memory (GDM) show a strong architectural resemblance as they are based on the Growing When Required (GWR) (Marsland et al, 2002) approach
Summary
Social robots that interact with humans in their everyday lives are exposed to a dynamic and challenging environment This dynamic environment provides continuous data streams (Parisi et al, 2019) that are potentially infinite and non-stationary (Ghesmoune et al, 2016; Wiwatcharakoses and Berrar, 2019). Information that was never seen before is frequently observed in a dynamic environment, requiring a conventional DL approach to be retrained on new and previous observations (Parisi et al, 2018, 2019) This can lead to infeasible memory requirements since all previously observed training samples need to be explicitly stored. LL is a long-standing challenge in machine learning due to catastrophic forgetting, which states that continually learning from non-stationary data distributions generally leads to a decrease in the performance of previously learned tasks (Parisi et al, 2019)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
