Research‎ > ‎

Event-based Vision

This is a topic that I got exposed to when I moved to Zurich to work at the Robotics and Perception Group (UZH).

Event cameras, such as the Dynamic Vision Sensor (DVS), are bio-inspired vision sensors that output pixel-level brightness changes instead of standard intensity frames. They offer significant advantages over standard cameras, namely a very high dynamic range, no motion blur, and a latency in the order of microseconds. However, because the output is composed of a sequence of asynchronous events rather than actual intensity images, traditional vision algorithms cannot be applied, so that new algorithms that exploit the high temporal resolution and the asynchronous nature of the sensor are required.


2017 Misha Mahowald Prize for Neuromorphic Engineering

Our research on event cameras for robotic applications wins the 2017 Misha Mahowald Prize! The award recognizes outstanding achievement in the field of neuromorphic engineering.

Misha Mahowald Prize 2017



We are organizing the Workshop on Event-based Vision at IEEE International Conference on Robotics and Automation (ICRA) 2017, Singapore. Join us for this event!


Continuous-Time Visual-Inertial Trajectory Estimation with Event Cameras

In this paper, we leverage a continuous-time framework to perform trajectory estimation by fusing visual data from a moving event camera with inertial data from an IMU. This framework allows direct integration of the asynchronous events with micro-second accuracy and the inertial measurements at high frequency. The pose trajectory is approximated by a smooth curve in the space of rigid-body motions using cubic splines. This formulation significantly reduces the number of variables in trajectory estimation problems. We evaluate our method on real data from several scenes and compare the results against ground truth from a motion-capture system. We show superior performance of the proposed technique compared to non-batch event-based algorithms. We also show that both the map orientation and scale can be recovered accurately by fusing events and inertial data. To the best of our knowledge, this is the first work on visual-inertial fusion with event cameras using a continuous-time framework.

References:

E. Mueggler, G. Gallego, H. Rebecq, D. Scaramuzza
Continuous-Time Visual-Inertial Trajectory Estimation with Event Cameras
(Under review), 2017.
PDF


Accurate Angular Velocity Estimation with an Event Camera

Motion estimation by contrast maximization


We present an algorithm to estimate the rotational motion of an event camera. In contrast to traditional cameras, which produce images at a fixed rate, event cameras have independent pixels that respond asynchronously to brightness changes, with microsecond resolution. Our method leverages the type of information conveyed by these novel sensors (that is, edges) to directly estimate the angular velocity of the camera, without requiring optical flow or image intensity estimation. The core of the method is a contrast maximization design. The method performs favorably against ground truth data and gyroscopic measurements from an Inertial Measurement Unit, even in the presence of very high-speed motions (close to 1000 deg/s).

References:
G. Gallego and D. Scaramuzza
Accurate Angular Velocity Estimation with an Event Camera
IEEE Robotics and Automation Letters (RA-L), vol 2, no. 2, pp. 632-639, Apr. 2017.
doi,   PDF


EVO: Event-based, 6-DOF Parallel Tracking and Mapping in Real-Time

EVO: Event-based Visual Odometry


We present EVO, an Event-based Visual Odometry algorithm. Our algorithm successfully leverages the outstanding properties of event cameras to track fast camera motions while recovering a semi-dense 3D map of the environment. The implementation runs in real-time on a standard CPU and outputs up to several hundred pose estimates per second. Due to the nature of event cameras, our algorithm is unaffected by motion blur and operates very well in challenging, high dynamic range conditions with strong illumination changes. To achieve this, we combine a novel, event-based tracking approach based on image-to-model alignment with a recent event-based 3D reconstruction algorithm in a parallel fashion. Additionally, we show that the output of our pipeline can be used to reconstruct intensity images from the binary event stream, though our algorithm does not require such intensity information. We believe that this work makes significant progress in SLAM by unlocking the potential of event cameras. This allows us to tackle challenging scenarios that are currently inaccessible to standard cameras.

References:
EVO
H. Rebecq, T. Horstschaefer, G. Gallego, D. Scaramuzza
EVO: A Geometric Approach to Event-based 6-DOF Parallel Tracking and Mapping in Real-time
IEEE Robotics and Automation Letters (RA-L), vol2, no. 2, pp. 593-600, Apr. 2017.
doi,   PDF


The Event Camera Dataset and Simulator:
Event-based Data for Pose Estimation, Visual Odometry, and SLAM


We present the world's first collection of datasets with an event-based camera for high-speed robotics. The data also include intensity images, inertial measurements, and ground truth from a motion-capture system. An event-based camera is a revolutionary vision sensor with three key advantages: a measurement rate that is several orders of magnitude faster than standard cameras, a latency of microseconds, and a high dynamic range of 130 decibels. These properties enable the design of a new class of algorithms for high-speed robotics, where standard cameras suffer from motion blur and high latency. All the data are released both as text files and binary (i.e., rosbag) files. Find out more on the dataset website!

References:
http://rpg.ifi.uzh.ch/davis_data.html
E. Mueggler, H. Rebecq, G. Gallego, T. Delbruck, D. Scaramuzza
The Event-Camera Dataset and Simulator: Event-based Data for Pose Estimation, Visual Odometry, and SLAM
International Journal of Robotics Research (IJRR), Feb. 2017.
doi,   PDF    Dataset page


EMVS: Event-based Multi-View Stereo

3D reconstruction with a single event camera


We introduce the problem of Event-based Multi-View Stereo (EMVS) for event cameras and propose a solution to it. Unlike traditional MVS methods, which address the problem of estimating dense 3D structure from a set of known viewpoints, EMVS estimates semi-dense 3D structure from an event camera with known trajectory. Our EMVS solution elegantly exploits two inherent properties of an event camera: (i) its ability to respond to scene edges --which naturally provide semi-dense geometric information without any preprocessing operation-- and (ii) the fact that it provides continuous measurements as the sensor moves. Despite its simplicity (it can be implemented in a few lines of code), our algorithm is able to produce accurate, semi-dense depth maps. We successfully validate our method on both synthetic and real data. Our method is computationally very efficient and runs in real-time on a laptop CPU and even on a smartphone processor.

References:
H. Rebecq, G. Gallego, D. Scaramuzza
EMVS: Event-based Multi-View Stereo
British Machine Vision Conference (BMVC), York, UK, Sep. 19-22, 2016.
Best Industry Paper Award (sponsored by Nvidia and BMVA)
Oral Talk: acceptance rate 7%
Proceedings,   PDF



Event-based, 6-DOF Camera Tracking for High-Speed Applications

Ego-motion estimation with an event camera


We present an event-based approach for ego-motion estimation, which provides pose updates upon the arrival of each event, thus virtually eliminating latency. Our method is the first work addressing and demonstrating event-based pose tracking in six degrees-of-freedom (DOF) motions in realistic and natural scenes, and it is able to track high-speed motions. The method is successfully evaluated in both indoor and outdoor scenes.

References:

G. Gallego, Jon E. A. Lund, E. Mueggler, H. Rebecq., T. Delbruck, D. Scaramuzza
Event-based, 6-DOF Camera Tracking for High-Speed Applications
(Under review)

PDF


Low-Latency Visual Odometry using Event-based Feature Tracks

IROS 2016


EBCCSP 2016


We develop an event-based feature tracking algorithm for the DAVIS sensor and show how to integrate it in an event-based visual odometry pipeline. Features are first detected in the grayscale frames and then tracked asynchronously using the stream of events. The features are then fed to an event-based visual odometry pipeline that tightly interleaves robust pose optimization and probabilistic mapping. We show that our method successfully tracks the 6-DOF motion of the sensor in natural scenes (see video above).

References:
B. Kueng, E. Mueggler, G. Gallego, D. Scaramuzza
Low-Latency Visual Odometry using Event-based Feature Tracks
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Daejeon, Korea, Oct. 9-14, 2016.
Best Application Paper Award Finalist.
Highlight talk: acceptance rate 2.5%

doi,   PDF

D. Tedaldi, G. Gallego, E. Mueggler, D. Scaramuzza
Feature Detection and Tracking with the Dynamic and Active-pixel Vision Sensor (DAVIS)
IEEE International Conference on Event-Based Control, Communication, and Signal Processing (EBCCSP), Krakow, Poland, June 13-15, 2016.
doi,   PDF


Continuous-Time Trajectory Estimation for Event-based Vision Sensors


In this paper, we address ego-motion estimation for an event-based vision sensor using a continuous-time framework to directly integrate the information conveyed by the sensor. The DVS pose trajectory is approximated by a smooth curve in the space of rigid-body motions using cubic splines and it is optimized according to the observed events. We evaluate our method using datasets acquired from sensor-in-the-loop simulations and onboard a quadrotor performing flips. The results are compared to the ground truth, showing the good performance of the proposed technique.

References:
E. Mueggler, G. Gallego, D. Scaramuzza
Continuous-Time Trajectory Estimation for Event-based Vision Sensors
Robotics: Science and Systems XI (RSS), Rome, Italy, July 13-17, 2015.
doi PDF


Event-based Camera Pose Tracking using a Generative Event Model
We tackle the problem of event-based camera localization in a known environment, without additional sensing, using a probabilistic generative event model in a Bayesian filtering framework. Our main contribution is the design of the likelihood function used in the filter to process the observed events. Based on the physical characteristics of the sensor and on empirical evidence of the Gaussian-like distribution of spiked events with respect to the brightness change, we propose to use the contrast residual as a measure of how well the estimated pose of the event-based camera and the environment explain the observed events. The filter allows for localization in the general case of six degrees-of-freedom motions.

G. Gallego, C. Forster, E. Mueggler, D. Scaramuzza
Event-based Camera Pose Tracking using a Generative Event Model
arXiv:1510.01972, 2015.
PDF


Lifetime Estimation of Events from Dynamic Vision Sensors

We develop an algorithm that augments each event with its "lifetime", which is computed from the event's velocity on the image plane. The generated stream of augmented events gives a continuous representation of events in time, hence enabling the design of new algorithms that outperform those based on the accumulation of events over fixed, artificially-chosen time intervals. A direct application of this augmented stream is the construction of sharp gradient (edge-like) images at any time instant. We successfully demonstrate our method in different scenarios, including high-speed quadrotor flips, and compare it to standard visualization methods.

References:
E. Mueggler, C. Forster, N. Baumli, G. Gallego, D. Scaramuzza
Lifetime Estimation of Events from Dynamic Vision Sensors
IEEE International Conference on Robotics and Automation (ICRA), pp. 4874-4881, Seattle (WA), USA, May 26-30, 2015.
doi PDF