# GEN-SLAM: Generative Modeling for Monocular Simultaneous Localization and Mapping

## TL;DR

GEN-SLAM introduces a deep learning system that performs monocular localization and obstacle detection for mobile robots, learning from traditional SLAM methods to provide topological pose and depth maps using only a single camera.

## Contribution

It combines CNN-based topological localization with a conditional VAE for depth estimation, integrating geometric SLAM principles into a deep learning framework.

## Key findings

- Effective localization on simulated and real datasets
- Accurate depth estimation from monocular images
- Seamless integration of localization and obstacle avoidance

## Abstract

We present a Deep Learning based system for the twin tasks of localization and obstacle avoidance essential to any mobile robot. Our system learns from conventional geometric SLAM, and outputs, using a single camera, the topological pose of the camera in an environment, and the depth map of obstacles around it. We use a CNN to localize in a topological map, and a conditional VAE to output depth for a camera image, conditional on this topological location estimation. We demonstrate the effectiveness of our monocular localization and depth estimation system on simulated and real datasets.