Dataset: Central and Peripheral Human Vision in Collective Motion

Dataset: Central and Peripheral Human Vision in Collective Motion This dataset accompanies the study "Central and Peripheral Human Vision in Collective Motion" (Jarvis, Palle & Floreano, EPFL, 2026), which investigates the relative contributions of central and peripheral vision to coordinated human locomotion using an immersive virtual reality flight simulator. Experiment overview. 40 participants (recruited from EPFL, mean age 24.3 ± 3.0 years, 4 female) flew within a simulated flock of 60 birds using the Birdly™ flight simulator (Somniacs Ltd) paired with a Meta Quest Pro VR headset. Participants performed the task under three visual conditions: Normal Vision (full field of view), Peripheral Vision (central 10° radius masked), and Central Vision (periphery masked, only central 10° visible). Flight performance was quantified using the Reynolds Score, measuring adherence to the Reynolds flocking algorithm. Dataset structure. The dataset is organized into four folders: boid/ — Simulated flock data: positions of all 60 birds at each timestep across all flights normal_vision/ — Full field-of-view condition peripheral_vision/ — Central vision masked condition central_vision/ — Peripheral vision masked condition Each folder contains 40 .pkb files, one per participant, recording synchronized time-series data at approximately 8 Hz, including: timestamps, Birdly platform pose (pitch, roll, speed), VR headset orientation, binocular gaze direction, and the 3D positions of all birds in the flock. All data has been anonymised and contains no personally identifiable information. Funding. This work was supported by the Swiss National Science Foundation (grant no. 200020 212077). All experiments were approved by the Human Research Ethics Committee of École Polytechnique Fédérale de Lausanne.