DefVINS: Visual-Inertial Odometry for Deformable Scenes
A new deformable visual-inertial odometry framework that separates a rigid, IMU-anchored state from a non-rigid warp represented by an embedded deformation graph.
Electronic Engineer and researcher with international experience in Argentina, Spain, and Germany. Formerly in the oil industry, now completing a Ph.D. in Systems Engineering and Computer Science under Prof. Javier Civera’s supervision. Skilled in C++, Python, and MATLAB, with a strong focus on real-time systems, optimization, and modular design. Native speaker in Spanish, Fluent in English; conversational in Italian. Passionate about turning technology advances into robust, deployable solutions.
A new deformable visual-inertial odometry framework that separates a rigid, IMU-anchored state from a non-rigid warp represented by an embedded deformation graph.
A new analytical solution that is easy to implement and robust at initialization, thanks to the small-rotation and constant-velocity approximations, which simplify the problem while preserving the essential coupling between motion and inertial measurements.
A novel GNSS-inertial initialization strategy that delays the use of global GNSS measurements until sufficient information is available to accurately estimate the state of a sensorized device. A criterion based on the evolution of the Hessian matrix singular values is introduced.
A novel dataset designed to benchmark methods in the intersection between SLAM and novel view rendering. It consists of 40 sequences with synchronized RGB, depth, IMU, robot kinematic data, and ground-truth pose streams.
Estimate the camera motion in a rigid 3D environment, along with the state of an IMU while offering the flexibility to operate as a multiframe optimization or to marginalize older data.
Applying a new technique called compressive sensing in order to obtain the photocurrent map of photovoltaic devices without mechanical processes