Collection of papers in trajectory forecasting categorised according to the high-level structure
The literature survey is categorized as:
- Classical: Papers not utilizing neural networks for trajectory forecasting
- Motion-Based: Papers utilizing neural networks for trajectory forecasting without modelling interactions with neighbouring agents or physical spaces.
- Agent-Agent Interactions: Papers utilizing neural networks for trajectory forecasting modelling interactions with neighbouring agents but not physical spaces.
- Agent-Space Interactions: Papers utilizing neural networks for trajectory forecasting modelling interactions with physical spaces but not neighbouring agents.
- Agent-Agent-Space Interactions: Papers utilizing neural networks for trajectory forecasting modelling interactions with both physical spaces as well as neighbouring agents.
- Miscellaneous: Papers related to related topics like activity forecasting, human body dynamics
- Social Force Model for Pedestrian Dynamics, 1998 Paper
- Simulation of pedestrian dynamics using a two-dimensional cellular automaton, 2001 Paper
- Discrete Choice Models for Pedestrian Walking Behavior, 2006 Paper
- Continuum crowds, 2006 Paper
- Modelling Smooth Paths Using Gaussian Processes, 2007 Paper
- Reciprocal n-body Collision Avoidance (ORCA), 2008 Paper
- You’ll Never Walk Alone: Modeling Social Behavior for Multi-target Tracking, 2009 Paper
- Socially-Aware Large-Scale Crowd Forecasting, 2014 Paper
- Learning to Predict Trajectories of Cooperatively Navigating Agents, 2014 Paper
- Understanding pedestrian behaviors from stationary crowd groups, 2015 Paper
- Learning Social Etiquette: Human Trajectory Understanding In Crowded Scenes, 2016 Paper
- Point-based Path Prediction from Polar Histograms, 2016 Paper
- Bi-Prediction: Pedestrian Trajectory Prediction Based on Bidirectional LSTM Classification, 2017 Paper
- RED: A simple but effective Baseline Predictor for the TrajNet Benchmark, 2018 Paper
- Convolutional Neural Network for Trajectory Prediction, 2018 Paper
- Location-Velocity Attention for Pedestrian Trajectory Prediction, 2019 Paper
- The Simpler the Better: Constant Velocity for Pedestrian Motion Prediction, 2019 Paper
- Transformer Networks for Trajectory Forecasting, 2020 Paper
- Social LSTM: Human Trajectory Prediction in Crowded Spaces, 2016 Paper
- A Data-driven Model for Interaction-Aware Pedestrian Motion Prediction in Object Cluttered Environments, 2017 Paper
- Soft + Hardwired Attention: An LSTM Framework for Human Trajectory Prediction and Abnormal Event Detection, 2017 Paper
- Social Attention: Modeling Attention in Human Crowds, 2017 Paper
- 3DOF Pedestrian Trajectory Prediction Learned from Long-Term Autonomous Mobile Robot Deployment Data, 2017 Paper
- Encoding Crowd Interaction with Deep Neural Network for Pedestrian Trajectory Prediction, 2018 Paper
- Group LSTM: Group Trajectory Prediction in Crowded Scenarios, 2018 Paper
- MX-LSTM: mixing tracklets and vislets to jointly forecast trajectories and head poses, 2018 Paper
- StarNet: Pedestrian Trajectory Prediction using Deep Neural Network in Star Topology, 2019 Paper
- SR-LSTM: State Refinement for LSTM towards Pedestrian Trajectory Prediction, 2019 Paper
- Recursive Social Behavior Graph for Trajectory Prediction, 2020 Paper
- Collaborative Motion Prediction via Neural Motion Message Passing Paper
- Social GAN: Socially Acceptable Trajectories with Generative Adversarial Networks, 2018 Paper
- Social Ways: Learning Multi-Modal Distributions of Pedestrian Trajectories with GANs, 2019 Paper
- Which Way Are You Going? Imitative Decision Learning for Path Forecasting in Dynamic Scenes, 2019 Paper
- Analyzing the Variety Loss in the Context of Probabilistic Trajectory Prediction, 2019 Paper
- The Trajectron: Probabilistic Multi-Agent Trajectory Modeling With Dynamic Spatiotemporal Graphs, 2019 Paper
- STGAT: Modeling Spatial-Temporal Interactions for Human Trajectory Prediction, 2019 Paper
- Stochastic Trajectory Prediction with Social Graph Network, 2019 Paper
- Social-STGCNN: A Social Spatio-Temporal Graph Convolutional Neural Network for Human Trajectory Prediction, 2020 Paper
- It Is Not the Journey but the Destination: Endpoint Conditioned Trajectory Prediction, 2020 Paper
- STAR: Spatio-Temporal Graph Transformer Networks for Pedestrian Trajectory Prediction, 2020 Paper
- Context-Aware Trajectory Prediction in Crowded Spaces, 2017 Paper
- Human Trajectory Prediction using Spatially aware Deep Attention Models, 2017 Paper
- SS-LSTM: A Hierarchical LSTM Model for Pedestrian Trajectory Prediction, 2018 Paper
- A Data-driven Model for Interaction-aware Pedestrian Motion Prediction in Object Cluttered Environments, 2018 Paper
- Multi-Agent Tensor Fusion for Contextual Trajectory Prediction, 2019 Paper
- DESIRE: Distant Future Prediction in Dynamic Scenes with Interacting Agents, 2017 Paper
- SoPhie: An Attentive GAN for Predicting Paths Compliant to Social and Physical Constraints, 2019 Paper
- Peeking into the Future: Predicting Future Person Activities and Locations in Videos, 2019 Paper
- Social-BiGAT: Multimodal Trajectory Forecasting using Bicycle-GAN and Graph Attention Networks, 2019 Paper
- Social-WaGDAT: Interaction-aware Trajectory Prediction via Wasserstein Graph Double-Attention Network, 2020 Paper
- Trajectron++: Multi-Agent Generative Trajectory Forecasting With Heterogeneous Data for Control, 2020 Paper
- Reciprocal Learning Networks for Human Trajectory Prediction, 2020 Paper
- The Garden of Forking Paths: Towards Multi-Future Trajectory Prediction, 2020 Paper
- Dynamic and Static Context-aware LSTM for Multi-agent Motion Prediction, 2020 Paper
- Accurate and Diverse Sampling of Sequences based on a “Best of Many” Sample Objective, 2018 Paper
- Overcoming Limitations of Mixture Density Networks: A Sampling and Fitting Framework for Multimodal Future Prediction, 2019 Paper
- Scene Compliant Trajectory Forecast with Agent-Centric Spatio-Temporal Grids, 2020 Paper
- Goal-GAN: Multimodal Trajectory Prediction Based on Goal Position Estimation, 2020 Paper
- Trajectory Learning for Activity Understanding: Unsupervised, Multilevel, and Long-Term Adaptive Approach, 2011 Paper
- Activity forecasting, 2012 Paper
- Context-Based Pedestrian Path Prediction, 2014 Paper
- Pedestrian’s Trajectory Forecast in Public Traffic with Artificial Neural Networks, 2014 Paper
- Learning Intentions for Improved Human Motion Prediction, 2014 Paper
- Pedestrian Path, Pose, and Intention Prediction Through Gaussian Process Dynamical Models and Pedestrian Activity Recognition, 2019 Paper
- Gaussian Process Dynamical Models for Human Motion, 2008 Paper
- Recurrent Network Models for Human Dynamics, 2015 Paper
Comparison of popular human trajectory forecasting papers based on the datasets on which the methods have been evaluated.
Method | ETH/UCY | SDD | TrajNet++ | Multipath |
---|---|---|---|---|
S-LSTM | ✓ | |||
DESIRE | ✓ | |||
S-GAN | ✓ | |||
Sophie | ✓ | ✓ | ||
Trajectron | ✓ | |||
Social-BiGAT | ✓ | |||
Social-STGCNN | ✓ | |||
Multiverse | ✓ | |||
PECNet | ✓ | ✓ | ✓ | |
D-LSTM | ✓ | |||
Social-NCE | ✓ |
Evaluation on TrajNet++ is preferred in comparison to ETH/UCY as the test set and the evaluation protocol for TrajNet++ is fixed (and extensive!). More details here. The variation in ADE/FDE greatly reduces among different methods when evaluated on equal grounds on TrajNet++ (leaderboard) in comparison to the numbers reported on ETH/UCY.
If you are new to trajectory forecasting, do check out the TrajNet++ framework! TrajNet++ is a code-base with specific focus on human trajectory forecasting, and having more than 10 trajectory forecastng baselines already implemented.