OpenTraj

Human Trajectory Prediction Dataset Benchmark

We introduce existing datasets for Human Trajectory Prediction (HTP) task, and also provide tools to load, visualize and analyze datasets. So far multiple datasets are supported.

Publicly Available Datasets

Name	Description	Ref
ETH	2 top view scenes containing walking pedestrians `#Traj:[Peds=750]` `Coord=world-2D` `FPS=2.5`	website paper
UCY	3 scenes (Zara/Arxiepiskopi/University). Zara and University close to top view. Arxiepiskopi more inclined. `#Traj:[Peds=786]` `Coord=world-2D` `FPS=2.5`	website paper
PETS 2009	different crowd activities `#Traj:[?]` `Coord=image-2D` `FPS=7`	website paper
SDD	8 top view scenes recorded by drone contains various types of agents `#Traj:[Bikes=4210 Peds=5232 Skates=292 Carts=174 Cars=316 Buss=76 Total=10,300]` `Coord=image-2D` `FPS=30`	website paper dropbox
GC	Grand Central Train Station Dataset: 1 scene of 33:20 minutes of crowd trajectories `#Traj:[Peds=12,684]` `Coord=image-2D` `FPS=25`	dropbox paper
HERMES	Controlled Experiments of Pedestrian Dynamics (Unidirectional and bidirectional flows) `#Traj:[?]` `Coord=world-2D` `FPS=16`	website data
Waymo	High-resolution sensor data collected by Waymo self-driving cars `#Traj:[?]` `Coord=2D and 3D` `FPS=?`	website github
KITTI	6 hours of traffic scenarios. various sensors `#Traj:[?]` `Coord=image-3D + Calib` `FPS=10`	website
inD	Naturalistic Trajectories of Vehicles and Vulnerable Road Users Recorded at German Intersections `#Traj:[Total=11,500]` `Coord=world-2D` `FPS=25`	website paper
L-CAS	Multisensor People Dataset Collected by a Pioneer 3-AT robot `#Traj:[?]` `Coord=0` `FPS=0`	website
Edinburgh	People walking through the Informatics Forum (University of Edinburgh) `#Traj:[ped=+92,000]` `FPS=0`	website
Town Center	CCTV video of pedestrians in a busy downtown area in Oxford `#Traj:[peds=2,200]` `Coord=0` `FPS=0`	website
Wild Track	surveillance video dataset of students recorded outside the ETH university main building in Zurich. `#Traj:[peds=1,200]`	website
ATC	92 days of pedestrian trajectories in a shopping center in Osaka, Japan `#Traj:[?]` `Coord=world-2D + Range data`	website
VIRAT	Natural scenes showing people performing normal actions `#Traj:[?]` `Coord=0` `FPS=0`	website
Forking Paths Garden	Multi-modal Synthetic dataset, created in CARLA (3D simulator) based on real world trajectory data, extrapolated by human annotators `#Traj:[?]`	website github paper
DUT	Natural Vehicle-Crowd Interactions in crowded university campus `#Traj:[Peds=1,739 vehicles=123 Total=1,862]` `Coord=world-2D` `FPS=23.98`	github paper
CITR	Fundamental Vehicle-Crowd Interaction scenarios in controlled experiments `#Traj:[Peds=340]` `Coord=world-2D` `FPS=29.97`	github paper
nuScenes	Large-scale Autonomous Driving dataset `#Traj:[peds=222,164 vehicles=662,856]` `Coord=World + 3D Range Data` `FPS=2`	website
VRU	consists of pedestrian and cyclist trajectories, recorded at an urban intersection using cameras and LiDARs `#Traj:[peds=1068 Bikes=464]` `Coord=World (Meter)` `FPS=25`	website
City Scapes	25,000 annotated images (Semantic/ Instance-wise/ Dense pixel annotations) `#Traj:[?]`	website
Argoverse	320 hours of Self-driving dataset `#Traj:[objects=11,052]` `Coord=3D` `FPS=10`	website
Ko-PER	Trajectories of People and vehicles at Urban Intersections (Laserscanner + Video) `#Traj:[peds=350]` `Coord=world-2D`	paper
TRAF	small dataset of dense and heterogeneous traffic videos in India (22 footages) `#Traj:[Cars=33 Bikes=20 Peds=11]` `Coord=image-2D` `FPS=10`	website gDrive paper
ETH-Person	Multi-Person Data Collected from Mobile Platforms	website

Human Trajectory Prediction Benchmarks

Trajnet: Trajectory Forecasting Challenge
Trajnet++: Trajectory Forecasting Challenge
MOT-Challenge: Multiple Object Tracking Benchmark
JackRabbot: Detection And Tracking Dataset and Benchmark

Toolkit

To download the toolkit, separately in a zip file click: here

1. Benchmarks

Using python files in benchmarking/indicators dir, you can generate the results of each of the indicators presented in the article. For more information about each of the scripts check the information in toolkit.

2. Loaders

Using python files in loaders dir, you can load a dataset into a dataset object, which uses Pandas data frames to store the data. It would be super easy to retrieve the trajectories, using different queries (by agent_id, timestamp, ...).

3. Visualization

A simple script is added play.py, and can be used to visualize a given dataset:

References: an awsome list of trajectory prediction references can be found here

Contributions: Have any idea to improve the code? Fork the project, update it and submit a merge request.

Feel free to open new issues.

If you find this work useful in your research, then please cite:

@inproceedings{amirian2020opentraj,
      title={OpenTraj: Assessing Prediction Complexity in Human Trajectories Datasets}, 
      author={Javad Amirian and Bingqing Zhang and Francisco Valente Castro and Juan Jose Baldelomar and Jean-Bernard Hayet and Julien Pettre},
      booktitle={Asian Conference on Computer Vision (ACCV)},
      number={CONF},      
      year={2020},
      organization={Springer}
}

abduallahmohamed/OpenTraj