# Single Level Feature-to-Feature Forecasting with Deformable Convolutions

## TL;DR

This paper introduces a novel feature-to-feature forecasting method for future semantic segmentation in driving scenarios, utilizing deformable convolutions to model diverse motion patterns and achieve state-of-the-art results.

## Contribution

It proposes a deformable convolution-based approach for feature forecasting that focuses on abstract features, improving prediction accuracy with minimal parameter increase.

## Key findings

- Outperforms regular and dilated convolution models
- Achieves state-of-the-art on Cityscapes for nine-step forecasting
- Minimal parameter increase compared to baseline models

## Abstract

Future anticipation is of vital importance in autonomous driving and other decision-making systems. We present a method to anticipate semantic segmentation of future frames in driving scenarios based on feature-to-feature forecasting. Our method is based on a semantic segmentation model without lateral connections within the upsampling path. Such design ensures that the forecasting addresses only the most abstract features on a very coarse resolution. We further propose to express feature-to-feature forecasting with deformable convolutions. This increases the modelling power due to being able to represent different motion patterns within a single feature map. Experiments show that our models with deformable convolutions outperform their regular and dilated counterparts while minimally increasing the number of parameters. Our method achieves state of the art performance on the Cityscapes validation set when forecasting nine timesteps into the future.