Using road topology to improve cyclist path prediction
Open Access
- 1 June 2017
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 289-296
Abstract
We learn motion models for cyclist path prediction on real-world tracks obtained from a moving vehicle, and propose to exploit the local road topology to obtain better predictive distributions. The tracks are extracted from the Tsinghua-Daimler Cyclist Benchmark for cyclist detection, and corrected for vehicle egomotion. Tracks are then spatially aligned to local curves and crossings in the road. We study a standard approach for path prediction in the literature based on Kalman Filters, as well as a mixture of specialized filters related to specific road orientations at junctions. Our experiments demonstrate an improved prediction accuracy (up to 20% on sharp turns) of mixing specialized motion models for canonical directions, and prior knowledge on the road topology. The new track data complements the existing video, disparity and annotation data of the original benchmark, and will be made publicly available.Keywords
This publication has 21 references indexed in Scilit:
- Surround vehicles trajectory analysis with recurrent neural networks2016
- Learning Social Etiquette: Human Trajectory Understanding In Crowded Scenes2016
- Social LSTM: Human Trajectory Prediction in Crowded Spaces2016
- Driver and pedestrian awareness-based collision risk analysis2016
- A Controlled Interactive Multiple Model Filter for Combined Pedestrian Intention Recognition and Path Prediction2015
- Mixture of Switching Linear Dynamics to Discover Behavior Patterns in Object TracksIEEE Transactions on Pattern Analysis and Machine Intelligence, 2015
- Context-Based Pedestrian Path Prediction2014
- Will the Pedestrian Cross? A Study on Pedestrian Path PredictionIEEE Transactions on Intelligent Transportation Systems, 2013
- Vision meets robotics: The KITTI datasetThe International Journal of Robotics Research, 2013
- Trajectory Learning for Activity Understanding: Unsupervised, Multilevel, and Long-Term Adaptive ApproachIEEE Transactions on Pattern Analysis and Machine Intelligence, 2011