Videos

A few videos issued from various research projects.

If first html tag is indented, and this include is called after a list, the html tag is considered a list element, and things go wrong. Leaving a hidden unindented line here
On this Page

A few videos issued from various research projects.

If first html tag is indented, and this include is called after a list, the html tag is considered a list element, and things go wrong. Leaving a hidden unindented line here

These videos are also shown within research projects, along with longer discussions of the research they illustrate.

Uncertainty-aware RL with VAE State Denoising for High-occlusion Tasks (2025)

If first html tag is indented, and this include is called after a list, the html tag is considered a list element, and things go wrong. Leaving a hidden unindented line here
Distributing sand homogeneously on a target surface, despite major visual occlusions incurred by the sand plume. More information. Please cite [1].

Robot Manipulation of Amorphous Materials (2025)

If first html tag is indented, and this include is called after a list, the html tag is considered a list element, and things go wrong. Leaving a hidden unindented line here
Autonomous sand grading. More information. Please cite [2].

Diffusion-based Grasp Planning with Local Geometric Features

If first html tag is indented, and this include is called after a list, the html tag is considered a list element, and things go wrong. Leaving a hidden unindented line here
If first html tag is indented, and this include is called after a list, the html tag is considered a list element, and things go wrong. Leaving a hidden unindented line here
There are two dominant approaches in modern robot grasp planning: dense prediction and sampling-based methods. Dense prediction calculates viable grasps across the robot’s view but is limited to predicting one grasp per voxel. Sampling-based methods, on the other hand, encode multi-modal grasp distributions, allowing for different grasp approaches at a point. However, these methods rely on a global latent representation, which struggles to represent the entire field of view, resulting in coarse grasps. To address this, we introduce Implicit Grasp Diffusion (IGD), which combines the strengths of both methods by using implicit neural representations to extract detailed local features and sampling grasps from diffusion models conditioned on these features. More information. Please cite [3].

Multimodal Hand Trajectory Anticipation and Shared Control (2024)

If first html tag is indented, and this include is called after a list, the html tag is considered a list element, and things go wrong. Leaving a hidden unindented line here
We present an intent estimator, dubbed the Robot Trajectron (RT), that produces a probabilistic representation of the robot’s anticipated trajectory based on its recent position, velocity and acceleration history. By taking arm dynamics into account, RT can capture the operator’s intent better than other SOTA models that only use the arm’s position, making it particularly well-suited to assist in tasks where the operator’s intent is susceptible to change. We derive a novel shared-control solution that combines RT’s predictive capacity to a representation of the locations of potential reaching targets. More information. Please cite [4].

Autonomous Climbing (2019)

Beautiful video of a field test we ran in Death Valley in December 2018:

If first html tag is indented, and this include is called after a list, the html tag is considered a list element, and things go wrong. Leaving a hidden unindented line here

(IEEE Spectrum coverage, YouTube)

Part-based Grasp Generalization (2015)

Learning prototypical parts by which objects are often grasped, to the end of grasping novel objects. Please cite [5].

If first html tag is indented, and this include is called after a list, the html tag is considered a list element, and things go wrong. Leaving a hidden unindented line here
Video illustrating the capabilities of a robot that learned how to grasp via the method illustrated in the video directly below.
If first html tag is indented, and this include is called after a list, the html tag is considered a list element, and things go wrong. Leaving a hidden unindented line here
Video illustrating the part learning process.

Grasping Using Tactile And Visual Data (2015)

The robot learns what it feels like to grasp an object. This way, it can abort a grasp that feels unstable before lifting up (and potentially breaking) the object. Please cite [6], [7]

If first html tag is indented, and this include is called after a list, the html tag is considered a list element, and things go wrong. Leaving a hidden unindented line here
Video illustrating pose- and touch-based grasp stability estimation.

Grasp Densities (2010)

Learning grasp affordance models through autonomous interaction. More information. Please cite [8].

If first html tag is indented, and this include is called after a list, the html tag is considered a list element, and things go wrong. Leaving a hidden unindented line here
Video illustrating pose- and touch-based grasp stability estimation.

References

  1. [1]
    AREPO: Uncertainty-Aware Robot Ensemble Learning Under Extreme Partial Observability.
    Yurui Du, Louis Hanut, Herman Bruyninckx, and Renaud Detry.
    IEEE Robotics and Automation Letters, 2025.
    DOI Bib PDF Video
  2. [2]
    Robotic Framework for Iterative and Adaptive Profile Grading of Sand.
    Louis Hanut, Yurui Du, Andrew Vande Moere, Renaud Detry, and Herman Bruyninckx.
    In IEEE International Conference on Robotics and Automation, 2025.
    Bib PDF Video
  3. [3]
    Implicit Grasp Diffusion: Bridging the Gap between Dense Prediction and Sampling-based Grasping.
    Pinhao Song, Pengteng Li, and Renaud Detry.
    In Conference on Robot Learning, 2024.
    Bib PDF Video
  4. [4]
    Robot Trajectron: Trajectory Prediction-based Shared Control for Robot Manipulation.
    Pinhao Song, Pengteng Li, Erwin Aertbelien, and Renaud Detry.
    In IEEE International Conference on Robotics and Automation, 2024.
    DOI Bib PDF Video
  5. [5]
    Learning a Dictionary of Prototypical Grasp-predicting Parts from Grasping Experience.
    Renaud Detry, Carl Henrik Ek, Marianna Madry, and Danica Kragic.
    In IEEE International Conference on Robotics and Automation, 2013.
    DOI Bib PDF Video
  6. [6]
    Learning Tactile Characterizations Of Object- And Pose-specific Grasps.
    Yasemin Bekiroglu, Renaud Detry, and Danica Kragic.
    In IEEE/RSJ International Conference on Intelligent Robots and Systems, 2011.
    DOI Bib PDF Video
  7. [7]
    Learning the Tactile Signatures of Prototypical Object Parts for Robust Part-based Grasping of Novel Objects.
    Emil Hyttinen, Danica Kragic, and Renaud Detry.
    In IEEE International Conference on Robotics and Automation, 2015.
    DOI Bib PDF
  8. [8]
    Learning Grasp Affordance Densities.
    R. Detry, D. Kraft, O. Kroemer, L. Bodenhagen, J. Peters, N. Krüger, and J. Piater.
    Paladyn. Journal of Behavioral Robotics, 2011.
    DOI Bib PDF