Sidekick Policy Learning for Active Visual Exploration

07/29/2018
by   Santhosh K. Ramakrishnan, et al.
6

We consider an active visual exploration scenario, where an agent must intelligently select its camera motions to efficiently reconstruct the full environment from only a limited set of narrow field-of-view glimpses. While the agent has full observability of the environment during training, it has only partial observability once deployed, being constrained by what portions it has seen and what camera motions are permissible. We introduce sidekick policy learning to capitalize on this imbalance of observability. The main idea is a preparatory learning phase that attempts simplified versions of the eventual exploration task, then guides the agent via reward shaping or initial policy supervision. To support interpretation of the resulting policies, we also develop a novel policy visualization technique. Results on active visual exploration tasks with 360 scenes and 3D objects show that sidekicks consistently improve performance and convergence rates over existing methods. Code, data and demos are available.

READ FULL TEXT

page 8

page 13

page 14

page 24

page 25

page 26

research
10/14/2022

Learning Active Camera for Multi-Object Navigation

Getting robots to navigate to multiple objects autonomously is essential...
research
09/23/2019

Where to Look Next: Unsupervised Active Visual Exploration on 360° Input

We address the problem of active visual exploration of large 360 inputs....
research
06/11/2020

Exploration by Maximizing Rényi Entropy for Zero-Shot Meta RL

Exploring the transition dynamics is essential to the success of reinfor...
research
02/22/2022

Coordinate-Aligned Multi-Camera Collaboration for Active Multi-Object Tracking

Active Multi-Object Tracking (AMOT) is a task where cameras are controll...
research
11/10/2020

Perturbation-based exploration methods in deep reinforcement learning

Recent research on structured exploration placed emphasis on identifying...
research
08/21/2020

Learning Affordance Landscapes forInteraction Exploration in 3D Environments

Embodied agents operating in human spaces must be able to master how the...
research
12/28/2021

Embodied Learning for Lifelong Visual Perception

We study lifelong visual perception in an embodied setup, where we devel...

Please sign up or login with your details

Forgot password? Click here to reset