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

arXiv: https://arxiv.org/abs/1907.11475 · HTML: https://mlsanity.com/paper/1907.11475/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 26 Jul 2019.

## Built on by (most cited first, 10 of 10)

- [DINO-Foresight: Looking into the Future with DINO](https://mlsanity.com/paper/2412.11673/) (2024)
- [Real-time Object Detection for Streaming Perception](https://mlsanity.com/paper/2203.12338/) (2022)
- [Long-term Video Frame Interpolation via Feature Propagation](https://mlsanity.com/paper/2203.15427/) (2022)
- [Advancing Semantic Future Prediction through Multimodal Visual Sequence Transformers](https://mlsanity.com/paper/2501.08303/) (2025)
- [Panoptic Segmentation Forecasting](https://mlsanity.com/paper/2104.03962/) (2021)
- [StreamYOLO: Real-time Object Detection for Streaming Perception](https://mlsanity.com/paper/2207.10433/) (2022)
- [TCSloT: Text Guided 3D Context and Slope Aware Triple Network for Dental Implant Position Prediction](https://mlsanity.com/paper/2308.05355/) (2023)
- [Forecasting Future Instance Segmentation with Learned Optical Flow and Warping](https://mlsanity.com/paper/2211.08049/) (2022)
- [Dense Semantic Forecasting in Video by Joint Regression of Features and Feature Motion](https://mlsanity.com/paper/2101.10777/) (2021)
- [Joint Forecasting of Panoptic Segmentations with Difference Attention](https://mlsanity.com/paper/2204.07157/) (2022)

## Abstract (arXiv)

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.