We present a novel deep neural network architecture for representing robot experiences in an episodic-like memory that facilitates encoding, recalling, and predicting action experiences. Our proposed unsupervised deep episodic memory model as follows: First, encodes observed actions in a latent vector space and, based on this latent encoding, second, infers most similar episodes previously experienced, third, reconstructs original episodes, and finally, predicts future frames in an end-to-end fashion. Results show that conceptually similar actions are mapped into the same region of the latent vector space. Based on these results, we introduce an action matching and retrieval mechanism, benchmark its performance on two large-scale action datasets, 20BN-something-something and ActivityNet and evaluate its generalization capability in a real-world scenario on a humanoid robot.


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Deep Episodic Memory : Encoding, Recalling, and Predicting Episodic Experiences for Robot Action Execution


    Beteiligte:
    Rothfuss, Jonas (Autor:in) / Ferreira, Fabio (Autor:in) / Aksoy, Eren (Autor:in) / Zhou, You (Autor:in) / Asfour, Tamim (Autor:in)

    Erscheinungsdatum :

    2018-01-01


    Anmerkungen:

    Scopus 2-s2.0-85062299146



    Medientyp :

    Aufsatz (Zeitschrift)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Klassifikation :

    DDC:    006 / 629



    Tasking System for an Episodic Spacecraft

    Dragonette, R. / Bowman, A. | British Library Conference Proceedings | 2000


    Enhancing human computer interaction with episodic memory in a virtual guide

    Rabe, Felix / Wachsmuth, Ipke / Kurosu, Masaaki | BASE | 2013

    Freier Zugriff

    Episodic cosmic-ray modulation in the heliosphere

    Le Roux, J.A. / Potgieter, M.S. | Elsevier | 1989


    SCALABLE AND EFFICIENT EPISODIC MEMORY IN COGNITIVE PROCESSING FOR AUTOMATED SYSTEMS

    BHATTACHARYYA RAJAN / CHO YOUNGKWAN / KWON HYUKSEONG | Europäisches Patentamt | 2020

    Freier Zugriff

    Cognitively-inspired episodic imagination for self-driving vehicles

    Mahmoud, Sara / Svensson, Henrik / Thill, Serge | BASE | 2019

    Freier Zugriff