# MAVNet: an Effective Semantic Segmentation Micro-Network for MAV-based Tasks

## TL;DR

MAVNet is a compact, efficient neural network designed for real-time semantic segmentation on micro aerial vehicles, balancing speed and accuracy while being suitable for SWaP-constrained platforms, and supported by new datasets.

## Contribution

We introduce MAVNet, a lightweight neural network for real-time segmentation on MAVs, and provide two novel datasets for MAV tracking and inspection tasks.

## Key findings

- MAVNet has 400 times fewer parameters than ERFNet.
- Achieves up to 48 FPS on NVIDIA 1080Ti and 9 FPS on Jetson Xavier.
- Performs comparably to larger models in empirical tests.

## Abstract

Real-time semantic image segmentation on platforms subject to size, weight and power (SWaP) constraints is a key area of interest for air surveillance and inspection. In this work, we propose MAVNet: a small, light-weight, deep neural network for real-time semantic segmentation on micro Aerial Vehicles (MAVs). MAVNet, inspired by ERFNet, features 400 times fewer parameters and achieves comparable performance with some reference models in empirical experiments. Our model achieves a trade-off between speed and accuracy, achieving up to 48 FPS on an NVIDIA 1080Ti and 9 FPS on the NVIDIA Jetson Xavier when processing high resolution imagery. Additionally, we provide two novel datasets that represent challenges in semantic segmentation for real-time MAV tracking and infrastructure inspection tasks and verify MAVNet on these datasets. Our algorithm and datasets are made publicly available.