no code implementations • 16 Apr 2024 • Ashkan Mirzaei, Riccardo de Lutio, Seung Wook Kim, David Acuna, Jonathan Kelly, Sanja Fidler, Igor Gilitschenski, Zan Gojcic
In this work, we propose an approach for 3D scene inpainting -- the task of coherently replacing parts of the reconstructed scene with desired content.
no code implementations • 19 Jan 2024 • Oliver Limoyo, Jimmy Li, Dmitriy Rivkin, Jonathan Kelly, Gregory Dudek
We leverage a visual language model (VLM) and an object detector to characterize the reference images via textual descriptions and then use a large language model (LLM) to retrieve relevant reference images based on a user's language query through text-based reasoning.
no code implementations • 4 Dec 2023 • Oliver Limoyo, Abhisek Konar, Trevor Ablett, Jonathan Kelly, Francois R. Hogan, Gregory Dudek
By doing so, the policy can generalize to object placement scenarios outside of the training environment without privileged information (e. g., placing a plate picked up from a table).
no code implementations • 2 Nov 2023 • Trevor Ablett, Oliver Limoyo, Adam Sigal, Affan Jilani, Jonathan Kelly, Kaleem Siddiqi, Francois Hogan, Gregory Dudek
An STS sensor can be switched between visual and tactile modes by leveraging a semi-transparent surface and controllable lighting, allowing for both pre-contact visual sensing and during-contact tactile sensing with a single sensor.
no code implementations • 27 Oct 2023 • Tristan Aumentado-Armstrong, Ashkan Mirzaei, Marcus A. Brubaker, Jonathan Kelly, Alex Levinshtein, Konstantinos G. Derpanis, Igor Gilitschenski
The resulting latent-space NeRF can produce novel views with higher quality than standard colour-space NeRFs, as the AE can correct certain visual artifacts, while rendering over three times faster.
no code implementations • 17 Aug 2023 • Ashkan Mirzaei, Tristan Aumentado-Armstrong, Marcus A. Brubaker, Jonathan Kelly, Alex Levinshtein, Konstantinos G. Derpanis, Igor Gilitschenski
A field is trained on relevance maps of training views, denoted as the relevance field, defining the 3D region within which modifications should be made.
no code implementations • 15 May 2023 • Sean Wu, Nicole Amenta, Jiachen Zhou, Sandro Papais, Jonathan Kelly
Following four successful years in the SAE AutoDrive Challenge Series I, the University of Toronto is participating in the Series II competition to develop a Level 4 autonomous passenger vehicle capable of handling various urban driving scenarios by 2025.
no code implementations • 7 May 2023 • Andrej Janda, Pierre Merriaux, Pierre Olivier, Jonathan Kelly
The shape of the return waveform is affected by many factors, including the distance that the light pulse travels and the angle of incidence with a surface.
no code implementations • ICCV 2023 • Ashkan Mirzaei, Tristan Aumentado-Armstrong, Marcus A. Brubaker, Jonathan Kelly, Alex Levinshtein, Konstantinos G. Derpanis, Igor Gilitschenski
The popularity of Neural Radiance Fields (NeRFs) for view synthesis has led to a desire for NeRF editing tools.
1 code implementation • 13 Feb 2023 • Philippe Nadeau, Matthew Giamou, Jonathan Kelly
To operate safely and efficiently alongside human workers, collaborative robots (cobots) require the ability to quickly understand the dynamics of manipulated objects.
no code implementations • 18 Jan 2023 • Andrej Janda, Brandon Wagstaff, Edwin G. Ng, Jonathan Kelly
Reducing the quantity of annotations required for supervised training is vital when labels are scarce and costly.
1 code implementation • 30 Dec 2022 • Trevor Ablett, Bryan Chan, Jonathan Kelly
In this work, we show that the standard, naive approach to exploration can manifest as a suboptimal local maximum if a policy learned with AIL sufficiently matches the expert distribution without fully learning the desired task.
no code implementations • CVPR 2023 • Ashkan Mirzaei, Tristan Aumentado-Armstrong, Konstantinos G. Derpanis, Jonathan Kelly, Marcus A. Brubaker, Igor Gilitschenski, Alex Levinshtein
We refer to this task as 3D inpainting.
no code implementations • 21 Nov 2022 • Andrej Janda, Brandon Wagstaff, Edwin G. Ng, Jonathan Kelly
Reducing the quantity of annotations required for supervised training is vital when labels are scarce and costly.
no code implementations • 4 Jul 2022 • Ashkan Mirzaei, Yash Kant, Jonathan Kelly, Igor Gilitschenski
In this paper we introduce LaTeRF, a method for extracting an object of interest from a scene given 2D images of the entire scene, known camera poses, a natural language description of the object, and a set of point-labels of object and non-object points in the input images.
1 code implementation • 21 Apr 2022 • Oliver Limoyo, Trevor Ablett, Jonathan Kelly
In this work, we present a self-supervised generative modelling framework to jointly learn a probabilistic latent state representation of multimodal data and the respective dynamics.
no code implementations • 19 Feb 2022 • Abhinav Grover, Philippe Nadeau, Christopher Grebe, Jonathan Kelly
The ability to perceive object slip via tactile feedback enables humans to accomplish complex manipulation tasks including maintaining a stable grasp.
1 code implementation • 16 Dec 2021 • Trevor Ablett, Bryan Chan, Jonathan Kelly
We present Learning from Guided Play (LfGP), a framework in which we leverage expert demonstrations of, in addition to a main task, multiple auxiliary tasks.
1 code implementation • 7 Jun 2021 • Brandon Wagstaff, Valentin Peretroukhin, Jonathan Kelly
Structure from motion (SfM) has recently been formulated as a self-supervised learning problem, where neural network models of depth and egomotion are learned jointly through view synthesis.
no code implementations • 1 Jun 2021 • Jonathan Kelly, Christopher Grebe, Matthew Giamou
We examine the problem of time delay estimation, or temporal calibration, in the context of multisensor data fusion.
1 code implementation • 28 Apr 2021 • Trevor Ablett, Yifan Zhai, Jonathan Kelly
In this work, we demonstrate that a multiview policy can be found through imitation learning by collecting data from a variety of viewpoints.
no code implementations • 24 Mar 2021 • Abhinav Grover, Christopher Grebe, Philippe Nadeau, Jonathan Kelly
Despite the utility of tactile information, tactile sensors have yet to be widely deployed in industrial robotics settings.
no code implementations • 8 Feb 2021 • Justin Tomasi, Brandon Wagstaff, Steven L. Waslander, Jonathan Kelly
Successful visual navigation depends upon capturing images that contain sufficient useful information.
no code implementations • 6 Oct 2020 • Alexis Morris, Hallie Siegel, Jonathan Kelly
Building trustworthy autonomous systems is challenging for many reasons beyond simply trying to engineer agents that 'always do the right thing.'
Computers and Society Human-Computer Interaction Robotics Software Engineering
1 code implementation • 1 Oct 2020 • Erik Hemberg, Jonathan Kelly, Michal Shlapentokh-Rothman, Bryn Reinstadler, Katherine Xu, Nick Rutar, Una-May O'Reilly
Many public sources of cyber threat and vulnerability information exist to serve the defense of cyber systems.
Cryptography and Security
1 code implementation • 8 Sep 2020 • Brandon Wagstaff, Jonathan Kelly
The self-supervised loss formulation for jointly training depth and egomotion neural networks with monocular images is well studied and has demonstrated state-of-the-art accuracy.
1 code implementation • 18 Aug 2020 • Oliver Limoyo, Bryan Chan, Filip Marić, Brandon Wagstaff, Rupam Mahmood, Jonathan Kelly
Learning or identifying dynamics from a sequence of high-dimensional observations is a difficult challenge in many domains, including reinforcement learning and control.
1 code implementation • 1 Jun 2020 • Valentin Peretroukhin, Matthew Giamou, David M. Rosen, W. Nicholas Greene, Nicholas Roy, Jonathan Kelly
Accurate rotation estimation is at the heart of robot perception tasks such as visual odometry and object pose estimation.
1 code implementation • 17 May 2020 • Emmett Wise, Matthew Giamou, Soroush Khoubyarian, Abhinav Grover, Jonathan Kelly
Correct fusion of data from two sensors is not possible without an accurate estimate of their relative pose, which can be determined through the process of extrinsic calibration.
Robotics
no code implementations • 7 Apr 2020 • Una-May O'Reilly, Jamal Toutouh, Marcos Pertierra, Daniel Prado Sanchez, Dennis Garcia, Anthony Erb Luogo, Jonathan Kelly, Erik Hemberg
We delineate Adversarial Genetic Programming for Cyber Security, a research topic that, by means of genetic programming (GP), replicates and studies the behavior of cyber adversaries and the dynamics of their engagements.
1 code implementation • 27 Feb 2020 • Brandon Wagstaff, Valentin Peretroukhin, Jonathan Kelly
We present a self-supervised deep pose correction (DPC) network that applies pose corrections to a visual odometry estimator to improve its accuracy.
Robotics
1 code implementation • 1 Oct 2019 • Brandon Wagstaff, Valentin Peretroukhin, Jonathan Kelly
We present two novel techniques for detecting zero-velocity events to improve foot-mounted inertial navigation.
Robotics Signal Processing
2 code implementations • 20 Sep 2019 • Filip Maric, Matthew Giamou, Soroush Khoubyarian, Ivan Petrovic, Jonathan Kelly
This is accomplished through a novel formulation of inverse kinematics as a nearest point problem, and with a fast sum of squares solver that exploits the sparsity of kinematic constraints for serial manipulators.
Robotics Optimization and Control
1 code implementation • 2 Apr 2019 • Matthew Giamou, Filip Maric, Valentin Peretroukhin, Jonathan Kelly
Estimating unknown rotations from noisy measurements is an important step in SfM and other 3D vision tasks.
1 code implementation • 1 Apr 2019 • Lee Clement, Mona Gridseth, Justin Tomasi, Jonathan Kelly
Long-term metric self-localization is an essential capability of autonomous mobile robots, but remains challenging for vision-based systems due to appearance changes caused by lighting, weather, or seasonal variations.
no code implementations • 1 Apr 2019 • Valentin Peretroukhin, Brandon Wagstaff, Matthew Giamou, Jonathan Kelly
Accurate estimates of rotation are crucial to vision-based motion estimation in augmented reality and robotics.
1 code implementation • 7 Oct 2018 • YiLun Wu, Xintong Du, Rikky Duivenvoorden, Jonathan Kelly
In particular, the details in this paper and the available open-source materials should enable learners to gain an understanding of aerodynamics, flight control, state estimation, software design, and simulation, while experimenting with a unique aerial robot.
Robotics
1 code implementation • 10 Sep 2018 • Matthew Giamou, Ziye Ma, Valentin Peretroukhin, Jonathan Kelly
We present a certifiably globally optimal algorithm for determining the extrinsic calibration between two sensors that are capable of producing independent egomotion estimates.
Robotics
1 code implementation • 13 Jul 2018 • Brandon Wagstaff, Jonathan Kelly
While existing threshold-based zero-velocity detectors are not robust to varying motion types, our learned model accurately detects stationary periods of the inertial measurement unit (IMU) despite changes in the motion of the user.
Robotics Human-Computer Interaction
1 code implementation • 10 Sep 2017 • Valentin Peretroukhin, Jonathan Kelly
We use this loss to train a Deep Pose Correction network (DPC-Net) that predicts corrections for a particular estimator, sensor and environment.
1 code implementation • 9 Sep 2017 • Lee Clement, Jonathan Kelly
Direct visual localization has recently enjoyed a resurgence in popularity with the increasing availability of cheap mobile computing power.
no code implementations • 1 Aug 2017 • Valentin Peretroukhin, William Vega-Brown, Nicholas Roy, Jonathan Kelly
Many algorithms in computer vision and robotics make strong assumptions about uncertainty, and rely on the validity of these assumptions to produce accurate and consistent state estimates.
no code implementations • 1 Aug 2017 • Valentin Peretroukhin, Lee Clement, Matthew Giamou, Jonathan Kelly
Navigation in unknown, chaotic environments continues to present a significant challenge for the robotics community.
1 code implementation • 4 Jul 2017 • Brandon Wagstaff, Valentin Peretroukhin, Jonathan Kelly
We present a method to improve the accuracy of a foot-mounted, zero-velocity-aided inertial navigation system (INS) by varying estimator parameters based on a real-time classification of motion type.
Robotics Human-Computer Interaction
2 code implementations • 20 Sep 2016 • Valentin Peretroukhin, Lee Clement, Jonathan Kelly
We present a method to incorporate global orientation information from the sun into a visual odometry pipeline using only the existing image stream, where the sun is typically not visible.
no code implementations • 15 Sep 2016 • Lee Clement, Valentin Peretroukhin, Jonathan Kelly
In the absence of reliable and accurate GPS, visual odometry (VO) has emerged as an effective means of estimating the egomotion of robotic vehicles.