# SplitNet: Sim2Sim and Task2Task Transfer for Embodied Visual Navigation

## TL;DR

SplitNet introduces a decoupled learning approach for embodied visual navigation, enhancing transferability between simulators and to real-world environments by explicitly separating perception and policy learning.

## Contribution

The paper presents SplitNet, a novel method that explicitly decomposes visual perception and policy learning, improving transferability and generalization in embodied navigation tasks.

## Key findings

- Significant improvements in simulator transfer performance.
- Enhanced generalization to unseen environments.
- Effective adaptation with limited target domain data.

## Abstract

We propose SplitNet, a method for decoupling visual perception and policy learning. By incorporating auxiliary tasks and selective learning of portions of the model, we explicitly decompose the learning objectives for visual navigation into perceiving the world and acting on that perception. We show dramatic improvements over baseline models on transferring between simulators, an encouraging step towards Sim2Real. Additionally, SplitNet generalizes better to unseen environments from the same simulator and transfers faster and more effectively to novel embodied navigation tasks. Further, given only a small sample from a target domain, SplitNet can match the performance of traditional end-to-end pipelines which receive the entire dataset. Code is available https://github.com/facebookresearch/splitnet