Self-Supervised Action-Space Prediction for Automated Driving

  title={Self-Supervised Action-Space Prediction for Automated Driving},
  author={Faris Janjos and Maxim Dolgov and Johann Marius Z{\"o}llner},
  journal={2021 IEEE Intelligent Vehicles Symposium (IV)},
Making informed driving decisions requires reliable prediction of other vehicles' trajectories. In this paper, we present a novel learned multi-modal trajectory prediction architecture for automated driving. It achieves kinematically feasible predictions by casting the learning problem into the space of accelerations and steering angles - by performing action-space prediction, we can leverage valuable model knowledge. Additionally, the dimensionality of the action manifold is lower than that of… 

Figures and Tables from this paper

Differentiable Constrained Imitation Learning for Robot Motion Planning and Control

This approach constitutes a general framework for constraint robotic motion planning and control using offline Imitation Learning methods in a differentiable manner, via explicit completion and gradient-based correction.

StarNet: Joint Action-Space Prediction with Star Graphs and Implicit Global-Frame Self-Attention

This work presents a novel multi-modal multi-agent trajectory prediction architecture, focusing on map and interaction modeling using graph representation, and extends this architecture to joint scene-level prediction, which produces multiple agents’ predictions simultaneously.



TNT: Target-driveN Trajectory Prediction

The key insight is that for prediction within a moderate time horizon, the future modes can be effectively captured by a set of target states, which leads to the target-driven trajectory prediction (TNT) framework.

DiversityGAN: Diversity-Aware Vehicle Motion Prediction via Latent Semantic Sampling

This work first extends the generative adversarial network framework with a low-dimensional approximate semantic space, and shape that space to capture semantics such as merging and turning, and sample from this space in a way that mimics the predicted distribution, but allows us to control coverage of semantically distinct outcomes.

Rules of the Road: Predicting Driving Behavior With a Convolutional Model of Semantic Interactions

A unified representation is presented which encodes such high-level semantic information in a spatial grid, allowing the use of deep convolutional models to fuse complex scene context and empirically show that one can effectively learn fundamentals of driving behavior.

What-If Motion Prediction for Autonomous Driving

This work proposes a recurrent graph-based attentional approach with interpretable geometric and social relationships that supports the injection of counterfactual geometric goals and social contexts that could be used in the planning loop to reason about unobserved causes or unlikely futures that are directly relevant to the AV's intended route.

Multimodal Trajectory Predictions for Autonomous Driving using Deep Convolutional Networks

This work presents a method to predict multiple possible trajectories of actors while also estimating their probabilities, and successfully tested on SDVs in closed-course tests.

Multiple Trajectory Prediction with Deep Temporal and Spatial Convolutional Neural Networks

A convolutional network that operates on rasterized actor-centric images which encode the static and dynamic actor-environment to make better use of the past movement of the actor and rely on uncertainties estimated from the previous object tracking stage is proposed.

Learning Interaction-Aware Probabilistic Driver Behavior Models from Urban Scenarios

This work proposes to learn Markovian action models with deep neural networks that are conditioned on a driver's route intention and the situational context, and demonstrates that the learned model outperforms the previous hand-tuned rule-based model.

MultiPath: Multiple Probabilistic Anchor Trajectory Hypotheses for Behavior Prediction

This work presents MultiPath, which leverages a fixed set of future state-sequence anchors that correspond to modes of the trajectory distribution, which is efficient, requiring only one forward inference pass to obtain multi-modal future distributions, and the output is parametric, allowing compact communication and analytical probabilistic queries.

Action-Based Representation Learning for Autonomous Driving

This work shows that an affordance-based driving model pre-trained with this approach can leverage a relatively small amount of weakly annotated imagery and outperform pure end-to-end driving models, while being more interpretable.

PnPNet: End-to-End Perception and Prediction With Tracking in the Loop

This work proposes PnPNet, an end-to-end model that takes as input sequential sensor data, and outputs at each time step object tracks and their future trajectories, and shows significant improvements over the state-of-the-art with better occlusion recovery and more accurate future prediction.