# Conditional Vehicle Trajectories Prediction in CARLA Urban Environment

arXiv: https://arxiv.org/abs/1909.00792 · HTML: https://mlsanity.com/paper/1909.00792/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Placed in Imitation and offline RL for driving, one of 13 directions in Offline RL and imitation learning from data.
- First posted on arXiv 2 Sep 2019.

## Placed in

- Offline RL and imitation learning from data / [Imitation and offline RL for driving](https://mlsanity.com/offlinerl/imitation-and-offline-rl-for-driving/)

## Built on by (most cited first, 6 of 6)

- [LanguageMPC: Large Language Models as Decision Makers for Autonomous Driving](https://mlsanity.com/paper/2310.03026/) (2023)
- [SMEMO: Social Memory for Trajectory Forecasting](https://mlsanity.com/paper/2203.12446/) (2022)
- [Adversarial Imitation Learning with Trajectorial Augmentation and Correction](https://mlsanity.com/paper/2103.13887/) (2021)
- [Data augmentation for efficient learning from parametric experts](https://mlsanity.com/paper/2205.11448/) (2022)
- [The Interstate-24 3D Dataset: a new benchmark for 3D multi-camera vehicle tracking](https://mlsanity.com/paper/2308.14833/) (2023)
- [A Reference Architecture of Reinforcement Learning Frameworks](https://mlsanity.com/paper/2603.06413/) (2026)

## Abstract (arXiv)

Imitation learning is becoming more and more successful for autonomous driving. End-to-end (raw signal to command) performs well on relatively simple tasks (lane keeping and navigation). Mid-to-mid (environment abstraction to mid-level trajectory representation) or direct perception (raw signal to performance) approaches strive to handle more complex, real life environment and tasks (e.g. complex intersection). In this work, we show that complex urban situations can be handled with raw signal input and mid-level representation. We build a hybrid end-to-mid approach predicting trajectories for neighbor vehicles and for the ego vehicle with a conditional navigation goal. We propose an original architecture inspired from social pooling LSTM taking low and mid level data as input and producing trajectories as polynomials of time. We introduce a label augmentation mechanism to get the level of generalization that is required to control a vehicle. The performance is evaluated on CARLA 0.8 benchmark, showing significant improvements over previously published state of the art.