Exploration of Optimized Semantic Segmentation Architectures for edge-Deployment on Drones

TL;DR

This paper analyzes and optimizes semantic segmentation architectures for drone applications, identifying the FPN-EfficientNetB3 model as the best balance of accuracy, memory, and speed, with significant improvements over previous models.

Contribution

It provides a comparative analysis of segmentation architectures for UAVs and identifies the optimal model with TensorRT optimizations for edge deployment.

Findings

FPN-EfficientNetB3 achieves IoU of 0.65 and F1-score of 0.71

Memory usage reduced by approximately 4.1 times

Inference latency improved by 10%

Abstract

In this paper, we present an analysis on the impact of network parameters for semantic segmentation architectures in context of UAV data processing. We present the analysis on the DroneDeploy Segmentation benchmark. Based on the comparative analysis we identify the optimal network architecture to be FPN-EfficientNetB3 with pretrained encoder backbones based on Imagenet Dataset. The network achieves IoU score of 0.65 and F1-score of 0.71 over the validation dataset. We also compare the various architectures in terms of their memory footprint and inference latency with further exploration of the impact of TensorRT based optimizations. We achieve memory savings of ~4.1x and latency improvement of 10% compared to Model: FPN and Backbone: InceptionResnetV2.