Yeah. That's me.

Sergio Caccamo

Robotics and Computer Vision Scientist

I am a Robotics and Computer Vision scientist at UNIVRSES in Stockholm where I mostly work on robotics and autonomous driving vision problems.

I did my PhD at the Robotics, Perception and Learning Lab (RPL) at KTH Royal Institute of Technology in Sweden. My main research area is robotics perception, specifically I was working on enhancing perception capabilities of field robots (Unmanned Ground Vehicles - UGVs ). My main supervisor is Danica Kragic and co-supervisor Petter Ögren. My research spans in the area of Perception, Computer Vision, Machine Learning and Control.

I spent 6 months as Intern at Mitsubishi Electric Research Labs (MERL) in Cambridge MA, US (Oct 2016 - May 2017). At MERL I worked on SLAM problems, objects reconstruction and Deep Learning applications.

During my PhD studies I was involved in the European project TRADR. In a typical TRADR scenario various kinds of field robots (UGVs and UAVs) explore a wide disaster area and collaborate with human teams (fire-fighters) over multiple asynchronous missions.
I was also Marie Curie Researcher (ESR) at Oslo University Hospital for 9 months working on image processing (organ segmentations) and I have been briefly involved in the European projects I3O and EMICAB.

“I, on the other hand, am a finished product. I absorb electrical energy directly and utilize it with an almost one hundred percent efficiency. I am composed of strong metal, am continuously conscious, and can stand extremes of environment easily.
These are facts which, with the self-evident proposition that no being can create another being superior to itself, smashes your silly hypothesis to nothing.”

- Isaac Asimov, I, Robot.

Publications & Patents

  • Enhancing Geometric Maps Through Environmental Interactions
    Sergio Caccamo
    KTH Doctoral Thesis 2018
  • Joint 3D Reconstruction of a Static Scene and Moving Objects
    Sergio Caccamo, Esra Ataer-Cansizoglu, Yuichi Taguchi
    3DV 2017
  • 3D Object Discovery and Modelling Using Single RGB-D Images Containing Multiple Object Instances
    Wim Abbeloos, Esra Ataer-Cansizoglu, Sergio Caccamo, Yuichi Taguchi, Yukiyasu Domae
    3DV 2017
  • RCAMP: Resilient Communication-Aware Motion Planner and Autonomous Repair of Wireless Connectivity in Mobile Robots
    Sergio Caccamo, Ramviyas Parasuraman, Luigi Freda, Mario Gianni, Petter Ogren
    IROS 2017
  • Detecting and Grouping Identical Objects for Region Proposal and Classification
    Sergio Caccamo* and Wim Abbeloos*, Esra Ataer-Cansizoglu, Yuichi Taguchi, Chen Feng, Teng-Yok Lee
    CVPR 2017 W: Deep Learning for Robotic Vision
  • An Approach to Retrieve from Communication Loss in Field Robots
    Ramviyas Parasuraman, Sergio Caccamo, Petter Ögren, Byung-Cheol Min
    RSS 2017 Workshop
  • Team KTH Amazon Picking Challenge 2016
    Team KTH
    ICRA 2017 Workshop
  • A New UGV Teleoperation Interface for Improved Awareness of Network Connectivity and Physical Surroundings
    Ramviyas Parasuraman, Sergio Caccamo, Petter Ögren, M.A.Neerincx
    HRI Journal
  • PATENT: SLAM algorithm for simultaneous mapping and object reconstruction
    Sergio Caccamo, Esra Ataer-Cansizoglu, Yuichi Taguchi
    In progress @ MERL. Cambridge MA, US.
  • PATENT: Algorithm for 3D Object Discovery and Modeling Using Single RGB-D Images
    Wim Abbeloos, Esra Ataer-Cansizoglu, Sergio Caccamo, Yuichi Taguchi, Yukiyasu Domae
    In progress @ MERL. Cambridge MA, US.
  • Active Perception and Modelling of Deformable Surfaces using Gaussian Processes and Position-based Dynamics
    S Caccamo, P Guler, H Kjellstrom, D Kragic
    Humanoids 2016
  • Active Exploration Using Gaussian Random Fields and Gaussian Process Implicit Surfaces
    S Caccamo, Y Bekiroglu, C H Ek, D Kragic
    IROS 2016
  • Free Look UGV Teleoperation Control Tested in Game Environment: Enhanced Performance and Reduced Workload
    F Båberg, S Caccamo, N Smets, M Neerincx, P Ögren
    SSRR 2016
  • Adaptive object centred teleoperation control of a mobile manipulator
    F Båberg, Y Wang, S Caccamo, P Ögren
    ICRA 2016
  • Extending a ugv teleoperation FLC interface with wireless network connectivity information
    S Caccamo, R Parasuraman, F Båberg, P Ögren
    IROS 2015
  • A computational model for motor learning in insects
    P Arena, S Caccamo, L Patané, R Strauss
    IJCNN 2013


A small list of main topics of interest and skills.

  • Coding:
    C++; Python; C; C#; Matlab; Java; ..
  • Tech:
    ROS; Matlab; Unity3D; UnrealE; PCL; OpenCV; Android; ..
  • Research areas:
    Perception; Machine learning; Computer Vision; Autonomous Driving; Robotics; Control;
  • Passions:
    Painting; Design; Coding; Video Games; Travelling; Football; Coffee; ..


Here below a list of research topics, publications and projects I have worked with. Description links might be still inactive.
*submitted/accepted to a journal or conference

Doctoral Thesis

Enhancing Geometric Maps Through Environmental Interactions.

Download the TRADR robot model for 3D printing

STL file. Dim 176 kB

Presentation at RosCon 2017

ROS for Search and Rescue Robotics: Tools and Lessons learned during TRADR.

Object reconstruction and mapping

Joint 3D Reconstruction of a Static Scene and Moving Objects.

Resilient communication-aware motion planner

Online WiFi mapping and self-reconnection strategies for mobile robots.

Active Perception of Deformable Surfaces

Enhancing environmental perception of non-rigid surfaces using Gaussian Processes.

Active Perception and Gaussian Processes

Enhancing environmental perception combining visual (RGB-D) and tactile data for object detection.

Robotics and Unity3D

Create large scale simulations of field robots using game engines (Unity3D).

Active Wifi perception on YouBot

Enhancing perception of wireless connectivity on field robots during teleoperation. IROS 2015

TRADR project

European FP7 project on collaborative search and rescue robots.

Image Processing for Organs Segmentation

Segmentation of liver, spleen and tumours using linear filtering on 3D structures.

Biological Systems and Machine Learning

A computational model for motor learning in insects using echo state recurrent networks. IJCNN 2013

Contact Me

PhD Sergio S. Caccamo.
Medborgarplatsen 3,
11428 Stockholm, Sweden


Recent (2017)

Previous (May 2017 ->)