Mariano Jaimez Tarifa
PhD. student in Computer Vision and Mobile Robotics


I am Mariano Jaimez Tarifa, a joint PhD student in the Machine Perception and Intelligent Robotics group (MAPIR) at the University of Málaga (Spain) and the Computer Vision group at the Technical University of Munich (Germany). My research interests include:

  • Computer Vision: Scene flow and visual odometry.
  • Mobile robotics: Autonomous navigation.
  • RGB-D cameras and their potential applications in the fields of robotics, computer vision and virtual/augmented reality.

Brief CV

I was born in Loja (Granada, Spain) in 1988. I received a B.Sc-M.Sc in "Ingeniería Industrial" (a very general engineering which covers mechanics, computer science, electronics, electricity, etc.) from the University of Málaga in 2010 with highest honor. I also got a M.Sc in Mechatronics in 2012. I received a grant (DPI2011-25483) from the National (Spanish) Plan of Research to do a 4-year PhD under the supervision of Prof. Javier González-Jiménez, which I started in January 2013. From March to July 2014 I was a guest researcher at the Computer Vision group of the Technical University of Munich, and in February 2015 I became a PhD student at the same University, pursuing a joint doctorate, under the supervision of Prof. Daniel Cremers.



Scene Flow: Scene flow is defined as the dense or semi-dense motion field of a scene observed at different instants of time. The potential applications of scene flow in the field of robotics are numerous: autonomous navigation and manipulation in dynamic environments, pose estimation or SLAM refinement, human-robot interaction or segmentation from motion are a few examples. Moreover, its usefulness goes beyond robotics and it can even be applied for human motion analysis and motion capture, virtual and augmented reality or driving assistance.

PD-Flow: In this work, me and my coauthors present the first dense real-time scene flow algorithm for RGB-D cameras. It imposes brightness constancy and geometric consistency of the solution, and is implemented on a GPU to achieve a high temporal performance.

MC-Flow: Here we address the problem of joint segmentation and motion estimation. We propose smooth piecewise segmentation that provides more realistic results than traditional sharp segmentations.

Code: PD-Flow (GitHub)

Code not available


Visual Odometry: Visual Odometry (VO) consists in the process of estimating the pose of an agent (typically a camera) from visual inputs. Nowadays, fast and accurate visual odometry is gaining importance in robotics respect to traditional solutions like wheel odometry or inertial navigation based on IMUs. In our work, we introduce a novel VO method that takes consecutive range images (or scans) to estimate the linear and angular velocity of the sensor. Our method only requires geometric data and, although it could theoretically work with any kind of range sensor, we have particularized its formulation to range cameras (DIFODO) and 2D laser scanners (RF2O).

For range cameras: It runs in real-time (30 Hz - 60 Hz) on a single CPU core. 

For 2D laser scanners: It needs less than 1 millisecond to be computed on a single CPU core. 

Code: Fast Visual Odometry for 3D range sensors (MRPT)

Code: Range Flow-based Visual 2D Odometry (ROS)


Reactive NavigationReactive navigation is a crucial component of nearly any mobile robot. It is one of two halves which, together with the path-planner, make up a navigation system according to the commonly used “hybrid architecture”. Within this scheme, a reactive navigator works at the low-level layer to guarantee safe and agile motions based on real-time sensor data. In our work we address the problem of planar navigation in indoor environments by a reactive navigation system which regards both the 3D shape of the robot and the 3D geometry of the environment. This navigator can be adapted to any wheeled robot, and has been extensively tested for years with different robotic platforms in varied and challenging scenarios. 

Code: 3D PTG-Based Reactive Navigator (MRPT) 






All Videos

Planar odometry from a Radial Laser Scanner (RF2O) (2016)

- Motion Cooperation: Smooth Piece-wise Scene Flow from RGB-D Images (2015)

- Fast Visual Odometry for 3-D range sensors (2014)

- Real-Time Dense Scene Flow for RGB-D Cameras (2014)

3D PTG-Based Reactive Navigation: overview (2013)

3D PTG-Based Reactive Navigation: Going through a tight clearance (2013)

Wheel-leg robotic module for hybrid locomotion (2011)

Omnibola (2010)



- Spherical robot OMNIBOLA©



Mariano Jaimez Tarifa 
Dpto. Ingenieria de Sistemas y Automatica
E.T.S.I. Informatica - Telecomunicacion
Universidad de Malaga

Campus Universitario de Teatinos
29071 Malaga, Spain
Phone: +34 952 13 3362
e-mail: marianojt [at]