Semantically-enhanced Deep Collision Prediction for Autonomous Navigation using Aerial Robots

TL;DR

This paper introduces a semantically-enhanced deep learning approach for aerial robots to predict collisions in cluttered environments with thin obstacles, without relying on maps or full pose data.

Contribution

It presents a novel modular method combining a semantically-enhanced Variational Autoencoder and an uncertainty-aware collision prediction network for improved navigation.

Findings

Effective in detecting thin obstacles in cluttered environments

Outperforms baseline methods in simulation and experiments

Enhances autonomous navigation safety and reliability

Abstract

This paper contributes a novel and modularized learning-based method for aerial robots navigating cluttered environments containing hard-to-perceive thin obstacles without assuming access to a map or the full pose estimation of the robot. The proposed solution builds upon a semantically-enhanced Variational Autoencoder that is trained with both real-world and simulated depth images to compress the input data, while preserving semantically-labeled thin obstacles and handling invalid pixels in the depth sensor's output. This compressed representation, in addition to the robot's partial state involving its linear/angular velocities and its attitude are then utilized to train an uncertainty-aware 3D Collision Prediction Network in simulation to predict collision scores for candidate action sequences in a predefined motion primitives library. A set of simulation and experimental studies in cluttered environments with various sizes and types of obstacles, including multiple hard-to-perceive thin objects, were conducted to evaluate the performance of the proposed method and compare against an end-to-end trained baseline. The results demonstrate the benefits of the proposed semantically-enhanced deep collision prediction for learning-based autonomous navigation.