mlsanityData to 7 Oct 2026

Towards Robust Human Trajectory Prediction in Raw Videos

arXiv 2108.08259 ↗First posted 18 Aug 2021cs.CV cs.ROcited by 7

Not on the map. Why?

Abstract (arXiv)

Human trajectory prediction has received increased attention lately due to its importance in applications such as autonomous vehicles and indoor robots. However, most existing methods make predictions based on human-labeled trajectories and ignore the errors and noises in detection and tracking. In this paper, we study the problem of human trajectory forecasting in raw videos, and show that the prediction accuracy can be severely affected by various types of tracking errors. Accordingly, we propose a simple yet effective strategy to correct the tracking failures by enforcing prediction consistency over time. The proposed "re-tracking" algorithm can be applied to any existing tracking and prediction pipelines. Experiments on public benchmark datasets demonstrate that the proposed method can improve both tracking and prediction performance in challenging real-world scenarios. The code and data are available at https://git.io/retracking-prediction.

Key facts. Cited by 7 papers in the map (data through 7 Oct 2026). Builds on 28 of its 42 references that are in the map. First posted on arXiv 18 Aug 2021.

Builds on

28 of 42 references.

References · 28

  1. 2017Mask R-CNNcited by 7,259
  2. 2016Simple Online and Realtime Trackingcited by 862
  3. 2018Social GAN: Socially Acceptable Trajectories with Generative Adversarial Networkscited by 687
  4. 2013Generating Sequences With Recurrent Neural Networkscited by 549
  5. 2020Social-STGCNN: A Social Spatio-Temporal Graph Convolutional Neural Network for Human Trajectory Predictioncited by 324
  6. 2018SoPhie: An Attentive GAN for Predicting Paths Compliant to Social and Physical Constraintscited by 296
  7. 2020It Is Not the Journey but the Destination: Endpoint Conditioned Trajectory Predictioncited by 235
  8. 2019Human Motion Trajectory Prediction: A Surveycited by 231
  9. 2019Social-BiGAT: Multimodal Trajectory Forecasting using Bicycle-GAN and Graph Attention Networkscited by 228
  10. 2020Fast and Furious: Real Time End-to-End 3D Detection, Tracking and Motion Forecasting with a Single Convolutional Netcited by 228
  11. 2020Spatio-Temporal Graph Transformer Networks for Pedestrian Trajectory Predictioncited by 218
  12. 2017Social Attention: Modeling Attention in Human Crowdscited by 185
16 more
  1. 2021IntentNet: Learning to Predict Intention from Raw Sensor Datacited by 158
  2. 2020PnPNet: End-to-End Perception and Prediction with Tracking in the Loopcited by 143
  3. 2015MOTChallenge 2015: Towards a Benchmark for Multi-Target Trackingcited by 133
  4. 2020Human Trajectory Forecasting in Crowds: A Deep Learning Perspectivecited by 128
  5. 2020Implicit Latent Variable Model for Scene-Consistent Motion Forecastingcited by 116
  6. 2019Peeking into the Future: Predicting Future Person Activities and Locations in Videoscited by 112
  7. 2019Spatially-Aware Graph Neural Networks for Relational Behavior Forecasting from Sensor Datacited by 104
  8. 2019What the Constant Velocity Model Can Teach Us About Pedestrian Motion Predictioncited by 78
  9. 2019Conditional Generative Neural System for Probabilistic Trajectory Predictioncited by 76
  10. 2017Long-Term On-Board Prediction of People in Traffic Scenes under Uncertaintycited by 63
  11. 2017Future Person Localization in First-Person Videoscited by 57
  12. 2020End-to-end Contextual Perception and Prediction with Interaction Transformercited by 54
  13. 2020LiRaNet: End-to-End Trajectory Prediction using Spatio-Temporal Radar Fusioncited by 18
  14. 2018MX-LSTM: mixing tracklets and vislets to jointly forecast trajectories and head posescited by 17
  15. 2018Tracking by Prediction: A Deep Generative Model for Mutli-Person localisation and Trackingcited by 9
  16. 2020AutoTrajectory: Label-free Trajectory Extraction and Prediction from Videos using Dynamic Pointscited by 6

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 1; 2022: 1; 2023: 1; 2024: 3; 2025: 0; 2026: 1.

7 papers.

  1. 2021MTP: Multi-Hypothesis Tracking and Prediction for Reduced Error Propagationcited by 14
  2. 2023DESTINE: Dynamic Goal Queries with Temporal Transductive Alignment for Trajectory Predictioncited by 5
  3. 2022T2FPV: Dataset and Method for Correcting First-Person View Errors in Pedestrian Trajectory Predictioncited by 4
  4. 2024StreamMOTP: Streaming and Unified Framework for Joint 3D Multi-Object Tracking and Trajectory Predictioncited by 3
  5. 2024DuoSpaceNet: Leveraging Both Bird's-Eye-View and Perspective View Representations for 3D Object Detectioncited by 2
  6. 2024RealTraj: Towards Real-World Pedestrian Trajectory Forecastingcited by 2
  7. 2026Uncertainty-Aware Trajectory Forecasting from Imperfect Trackingcited by 0