Social and Scene-Aware Trajectory Prediction in Crowded Spaces (2019-09-19T00:00:00.000000Z)

TL;DR

An LSTM (long short-term memory)-based model is constructed considering three fundamental factors: people interactions, past observations in terms of previously crossed areas and semantics of surrounding space, which corroborate the effectiveness of the proposed framework in comparison to L STM-based models for human path prediction.

Abstract

Mimicking human ability to forecast future positions or interpret complex interactions in urban scenarios, such as streets, shopping malls or squares, is essential to develop socially compliant robots or self-driving cars. Autonomous systems may gain advantage on anticipating human motion to avoid collisions or to naturally behave alongside people. To foresee plausible trajectories, we construct an LSTM (long short-term memory)-based model considering three fundamental factors: people interactions, past observations in terms of previously crossed areas and semantics of surrounding space. Our model encompasses several pooling mechanisms to join the above elements defining multiple tensors, namely social, navigation and semantic tensors. The network is tested in unstructured environments where complex paths emerge according to both internal (intentions) and external (other people, not accessible areas) motivations. As demonstrated, modeling paths unaware of social interactions or context information, is insufficient to correctly predict future positions. Experimental results corroborate the effectiveness of the proposed framework in comparison to LSTM-based models for human path prediction.

Authors

Matteo Lisotto

1 papers

Pasquale Coscia

1 papers

Lamberto Ballan

1 papers

Social and Scene-Aware Trajectory Prediction in Crowded Spaces

TL;DR

Abstract

Authors

References22 items

Group LSTM: Group Trajectory Prediction in Crowded Scenarios

RED: A Simple but Effective Baseline Predictor for the TrajNet Benchmark

Learning and Inferring “Dark Matter” and Predicting Human Intents and Trajectories in Videos

SoPhie: An Attentive GAN for Predicting Paths Compliant to Social and Physical Constraints

MX-LSTM: Mixing Tracklets and Vislets to Jointly Forecast Trajectories and Head Poses

Social GAN: Socially Acceptable Trajectories with Generative Adversarial Networks

Transferable Pedestrian Motion Prediction Models at Intersections

CAR-Net: Clairvoyant Attentive Recurrent Network

Social Attention: Modeling Attention in Human Crowds

Context-Aware Trajectory Prediction

Soft + Hardwired Attention: An LSTM Framework for Human Trajectory Prediction and Abnormal Event Detection

Learning Social Etiquette: Human Trajectory Understanding In Crowded Scenes

Social LSTM: Human Trajectory Prediction in Crowded Spaces

Knowledge Transfer for Scene-Specific Motion Prediction

Visualizing and Understanding Recurrent Networks

Context-Based Pedestrian Path Prediction

You'll never walk alone: Modeling social behavior for multi-target tracking

Crowds by Example

Social force model for pedestrian dynamics.

Long-term path prediction in urban scenarios using circular distributions

Trajnet: Towards a benchmark for human trajectory prediction

Research Showcase @ CMU

Field of Study

Journal Information

Name

Page