An open-source simulator for event cameras, providing accurate event generation with IMU and multi-camera support.
ESIM is an open-source event camera simulator that generates realistic event streams from 3D scenes for robotics and computer vision research. It simulates event-based vision sensors, IMU data, and provides ground truth information like depth maps and camera poses. The tool allows developers to test algorithms without physical event cameras.
Researchers and developers in robotics, computer vision, and event-based sensing who need to simulate event camera data for algorithm development and testing.
ESIM offers high accuracy through tight rendering integration, supports multi-camera and IMU simulation, and provides extensive ground truth data. Its open-source nature and ROS compatibility make it a flexible choice for event-based vision projects.
ESIM: an Open Event Camera Simulator
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Tight integration between rendering engine and simulator ensures realistic event streams, as emphasized in the CoRL paper for accuracy.
Provides camera poses, IMU biases, velocities, depth maps, and optic flow maps, enabling thorough validation of event-based algorithms.
Simulates IMU data and supports multi-camera systems, facilitating research on sensor fusion and complex perception setups.
Publishes to ROS and saves data to rosbag files, making it easy to integrate with existing robotics pipelines and tools.
Requires setup of UnrealCV, OpenGL, and assimp with installation instructions on a separate wiki, posing a significant barrier for new users.
Noise model is based on additive Gaussian noise on contrast thresholds and described as 'basic', which may not cover advanced real-world noise effects.
Python support is provided through the rpg_vid2e project, adding complexity for users who prefer integrated Python workflows over C++.