Mathieu Garon
YOU?
Author Swipe
View article: SpotLight: Shadow-Guided Object Relighting via Diffusion
SpotLight: Shadow-Guided Object Relighting via Diffusion Open
Recent work has shown that diffusion models can serve as powerful neural rendering engines that can be leveraged for inserting virtual objects into images. However, unlike typical physics-based renderers, these neural rendering engines are…
View article: ZeroComp: Zero-shot Object Compositing from Image Intrinsics via Diffusion
ZeroComp: Zero-shot Object Compositing from Image Intrinsics via Diffusion Open
We present ZeroComp, an effective zero-shot 3D object compositing approach that does not require paired composite-scene images during training. Our method leverages ControlNet to condition from intrinsic images and combines it with a Stabl…
View article: Learning to drive with neurological conditions: profile of users of an adapted driver training program and cognitive factors associated with success
Learning to drive with neurological conditions: profile of users of an adapted driver training program and cognitive factors associated with success Open
To describe the sociodemographic and cognitive profile of participants enrolled in an adapted driving program for individuals with neurological conditions, to explore the association between cognitive functioning and driving program outcom…
View article: Editable Indoor Lighting Estimation
Editable Indoor Lighting Estimation Open
We present a method for estimating lighting from a single perspective image of an indoor scene. Previous methods for predicting indoor illumination usually focus on either simple, parametric lighting that lack realism, or on richer represe…
View article: Deep Template-based Object Instance Detection
Deep Template-based Object Instance Detection Open
Much of the focus in the object detection literature has been on the problem of identifying the bounding box of a particular class of object in an image. Yet, in contexts such as robotics and augmented reality, it is often necessary to fin…
View article: Input Dropout for Spatially Aligned Modalities
Input Dropout for Spatially Aligned Modalities Open
Computer vision datasets containing multiple modalities such as color, depth, and thermal properties are now commonly accessible and useful for solving a wide array of challenging tasks. However, deploying multi-sensor heads is not possibl…
View article: RGB-D-E: Event Camera Calibration for Fast 6-DOF Object Tracking
RGB-D-E: Event Camera Calibration for Fast 6-DOF Object Tracking Open
Augmented reality devices require multiple sensors to perform various tasks such as localization and tracking. Currently, popular cameras are mostly frame-based (e.g. RGB and Depth) which impose a high data bandwidth and power usage. With …
View article: RGB-D-E: Event Camera Calibration for Fast 6-DOF Object Tracking
RGB-D-E: Event Camera Calibration for Fast 6-DOF Object Tracking Open
Augmented reality devices require multiple sensors to perform various tasks\nsuch as localization and tracking. Currently, popular cameras are mostly\nframe-based (e.g. RGB and Depth) which impose a high data bandwidth and power\nusage. Wi…
View article: Learning to Match Templates for Unseen Instance Detection
Learning to Match Templates for Unseen Instance Detection Open
Detecting objects in images is a quintessential problem in computer vision. Much of the focus in the literature has been on the problem of identifying the bounding box of a particular type of objects in an image. Yet, in many contexts such…
View article: Template-based Unseen Instance Detection
Template-based Unseen Instance Detection Open
Much of the focus in the object detection literature has been on the problem of identifying the bounding box of a particular class of object in an image. Yet, in contexts such as robotics and augmented reality, it is often necessary to fin…
View article: Fast Spatially-Varying Indoor Lighting Estimation
Fast Spatially-Varying Indoor Lighting Estimation Open
We propose a real-time method to estimate spatiallyvarying indoor lighting from a single RGB image. Given an image and a 2D location in that image, our CNN estimates a 5th order spherical harmonic representation of the lighting at the give…
View article: Deep 6-DOF Tracking
Deep 6-DOF Tracking Open
We present a temporal 6-DOF tracking method which leverages deep learning to achieve state-of-the-art performance on challenging datasets of real world capture. Our method is both more accurate and more robust to occlusions than the existi…