VUSFA:Variational Universal Successor Features Approximator to Improve Transfer DRL for Target Driven Visual Navigation

08/18/2019 ∙ by Shamane Siriwardhana, et al. ∙ 0 ∙ share

In this paper, we show how novel transfer reinforcement learning techniques can be applied to the complex task of target driven navigation using the photorealistic AI2THOR simulator. Specifically, we build on the concept of Universal Successor Features with an A3C agent. We introduce the novel architectural contribution of a Successor Feature Dependant Policy (SFDP) and adopt the concept of Variational Information Bottlenecks to achieve state of the art performance. VUSFA, our final architecture, is a straightforward approach that can be implemented using our open source repository. Our approach is generalizable, showed greater stability in training, and outperformed recent approaches in terms of transfer learning ability.

READ FULL TEXT

Visual Transfer for Reinforcement Learning via Wasserstein Domain Confusion

We introduce Wasserstein Adversarial Proximal Policy Optimization (WAPPO...
06/04/2020 ∙ by Josh Roy, et al. ∙ 12 ∙ share

read it
A Critical Investigation of Deep Reinforcement Learning for Navigation

The navigation problem is classically approached in two steps: an explor...
02/07/2018 ∙ by Vikas Dhiman, et al. ∙ 0 ∙ share

read it
AgentGraph: Towards Universal Dialogue Management with Structured Deep Reinforcement Learning

Dialogue policy plays an important role in task-oriented spoken dialogue...
05/27/2019 ∙ by Lu Chen, et al. ∙ 5 ∙ share

read it
Hierarchical Reinforcement Learning Framework towards Multi-agent Navigation

This paper proposes a navigation algorithm ori- ented to multi-agent dyn...
07/14/2018 ∙ by Wenhao Ding, et al. ∙ 0 ∙ share

read it
Transfer Learning for Related Reinforcement Learning Tasks via Image-to-Image Translation

Deep Reinforcement Learning has managed to achieve state-of-the-art resu...
05/31/2018 ∙ by Shani Gamrian, et al. ∙ 0 ∙ share

read it
Towards Target-Driven Visual Navigation in Indoor Scenes via Generative Imitation Learning

We present a target-driven navigation system to improve mapless visual n...
09/30/2020 ∙ by Qiaoyun Wu, et al. ∙ 0 ∙ share

read it
Target Driven Visual Navigation with Hybrid Asynchronous Universal Successor Representations

Being able to navigate to a target with minimal supervision and prior kn...
11/27/2018 ∙ by Shamane Siriwardhana, et al. ∙ 0 ∙ share

read it

VUSFA:Variational Universal Successor Features Approximator to Improve Transfer DRL for Target Driven Visual Navigation

Authors

Visual Transfer for Reinforcement Learning via Wasserstein Domain Confusion

A Critical Investigation of Deep Reinforcement Learning for Navigation

AgentGraph: Towards Universal Dialogue Management with Structured Deep Reinforcement Learning

Hierarchical Reinforcement Learning Framework towards Multi-agent Navigation

Transfer Learning for Related Reinforcement Learning Tasks via Image-to-Image Translation

Towards Target-Driven Visual Navigation in Indoor Scenes via Generative Imitation Learning

Target Driven Visual Navigation with Hybrid Asynchronous Universal Successor Representations

This week in AI