Visual Servoing in Autoencoder Latent Space: Supplementary material

Samuel Felton (1, 2) , Pascal Brault (2) , Elisa Fromont (1, 3) , Eric Marchand (2)

1 LACODAM - Large Scale Collaborative Data Mining
2 RAINBOW - Sensor-based and interactive robotics
3 IUF - Institut universitaire de France

This document provides additional information about [1], describing further experiments with Autoencoderbased visual servoing (AEVS), a method for visual servoing in the latent space of an autoencoder. Multiple experiments are described, both in a simulated environment and on a real robot. In this document, we first evaluate the impact of the network specification and dataset sizes on the visual servoing results. Then we conduct experiments to ascertain the applicability and transferability of the method to complex scenes.

Domaines: Robotique [cs.RO]

Samuel Felton, Pascal Brault, Elisa Fromont, Eric Marchand. Visual Servoing in Autoencoder Latent Space: Supplementary material. 2021. ⟨hal-03448667⟩