# Vision-Guided Outdoor Flight and Obstacle Evasion via Reinforcement Learning

arXiv: https://arxiv.org/abs/2605.24449 · HTML: https://mlsanity.com/paper/2605.24449/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in RL for robot navigation and UAV control, one of 15 directions in Deep reinforcement learning methods.
- Builds on 24 of its 35 references that are in the map.
- First posted on arXiv 23 May 2026.

## Placed in

- Deep reinforcement learning methods / [RL for robot navigation and UAV control](https://mlsanity.com/deeprl/rl-for-robot-navigation-and-uav-control/)

## Builds on (most cited first, 24 of 24)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [Learning High-Speed Flight in the Wild](https://mlsanity.com/paper/2110.05113/) (2021)
- [Flightmare: A Flexible Quadrotor Simulator](https://mlsanity.com/paper/2009.00563/) (2020)
- [DIML/CVL RGB-D Dataset: 2M RGB-D Images of Natural Indoor and Outdoor Scenes](https://mlsanity.com/paper/2110.11590/) (2021)
- [Autonomous Drone Racing: A Survey](https://mlsanity.com/paper/2301.01755/) (2023)
- [Generalization through Simulation: Integrating Simulated and Real Data into Deep Reinforcement Learning for Vision-Based Autonomous Flight](https://mlsanity.com/paper/1902.03701/) (2019)
- [Memory-based Deep Reinforcement Learning for Obstacle Avoidance in UAV with Limited Environment Knowledge](https://mlsanity.com/paper/1811.03307/) (2018)
- [Learning Perception-Aware Agile Flight in Cluttered Environments](https://mlsanity.com/paper/2210.01841/) (2022)
- [Computational Benefits of Intermediate Rewards for Goal-Reaching Policy Learning](https://mlsanity.com/paper/2107.03961/) (2021)
- [End-to-end Reinforcement Learning for Time-Optimal Quadcopter Flight](https://mlsanity.com/paper/2311.16948/) (2023)
- [A Learning-based Quadcopter Controller with Extreme Adaptation](https://mlsanity.com/paper/2409.12949/) (2024)
- [Search-based Motion Planning for Aggressive Flight in SE(3)](https://mlsanity.com/paper/1710.02748/) (2017)
- [Learning Deep Sensorimotor Policies for Vision-based Autonomous Drone Racing](https://mlsanity.com/paper/2210.14985/) (2022)
- [Aerial Gym Simulator: A Framework for Highly Parallelized Simulation of Aerial Robots](https://mlsanity.com/paper/2503.01471/) (2025)
- [Vision Transformers for End-to-End Vision-Based Quadrotor Obstacle Avoidance](https://mlsanity.com/paper/2405.10391/) (2024)
- [Reinforcement Learning for Collision-free Flight Exploiting Deep Collision Encoding](https://mlsanity.com/paper/2402.03947/) (2024)
- [MAVRL: Learn to Fly in Cluttered Environments with Varying Speed](https://mlsanity.com/paper/2402.08381/) (2024)
- [Large-scale Autonomous Flight with Real-time Semantic SLAM under Dense Forest Canopy](https://mlsanity.com/paper/2109.06479/) (2021)
- [Efficient Optical flow and Stereo Vision for Velocity Estimation and Obstacle Avoidance on an Autonomous Pocket Drone](https://mlsanity.com/paper/1612.06702/) (2016)
- [AvoidBench: A high-fidelity vision-based obstacle avoidance benchmarking suite for multi-rotors](https://mlsanity.com/paper/2301.07430/) (2023)
- [Rapid Quadrotor Navigation in Diverse Environments using an Onboard Depth Camera](https://mlsanity.com/paper/2411.04326/) (2024)
- [Rectangular Pyramid Partitioning using Integrated Depth Sensors (RAPPIDS): A Fast Planner for Multicopter Navigation](https://mlsanity.com/paper/2003.01245/) (2020)
- [Mapless-Planner: A Robust and Fast Planning Framework for Aggressive Autonomous Flight without Map Fusion](https://mlsanity.com/paper/2011.03975/) (2020)
- [Time-optimal Flight in Cluttered Environments via Safe Reinforcement Learning](https://mlsanity.com/paper/2406.19646/) (2024)

## Abstract (arXiv)

Although quadcopters boast impressive traversal capabilities enabled by their omnidirectional maneuverability, the need for continuous pilot control in complex environments impedes their application in GNSS and telemetry-denied scenarios. To this end, we propose a novel sensorimotor policy that uses stereo-vision depth and visual-inertial odometry (VIO) to autonomously navigate through obstacles in an unknown environment to reach a goal point. The policy is comprised of a pre-trained autoencoder as the perception head followed by a planning and control LSTM network which outputs velocity commands that can be followed by an off-the-shelf commercial drone. We leverage reinforcement and privileged learning paradigms to train the policy in simulation through a two-stage process: 1) initial training with optimal trajectories generated by a global motion planner acting as a supervisory backbone, 2) further fine-tuning in a curriculum environment. To bridge the sim-to-real gap, we employ domain randomization and reward shaping to create a policy that is both robust to noise and domain shift. In outdoor experiments, our approach achieves successful zero-shot transfer to both obstacle environments and a drone platform that were never encountered during training.