Showing 1–2 of 2 results for author: Prévost, D
-
Learning to Estimate Indoor Lighting from 3D Objects
Authors:
Henrique Weber,
Donald Prévost,
Jean-François Lalonde
Abstract:
In this work, we propose a step towards a more accurate prediction of the environment light given a single picture of a known object. To achieve this, we developed a deep learning method that is able to encode the latent space of indoor lighting using few parameters and that is trained on a database of environment maps. This latent space is then used to generate predictions of the light that are b…
▽ More
In this work, we propose a step towards a more accurate prediction of the environment light given a single picture of a known object. To achieve this, we developed a deep learning method that is able to encode the latent space of indoor lighting using few parameters and that is trained on a database of environment maps. This latent space is then used to generate predictions of the light that are both more realistic and accurate than previous methods. To achieve this, our first contribution is a deep autoencoder which is capable of learning the feature space that compactly models lighting. Our second contribution is a convolutional neural network that predicts the light from a single image of a known object. To train these networks, our third contribution is a novel dataset that contains 21,000 HDR indoor environment maps. The results indicate that the predictor can generate plausible lighting estimations even from diffuse objects.
△ Less
Submitted 13 August, 2018; v1 submitted 11 June, 2018;
originally announced June 2018.
-
Feed drive modelling for the simulation of tool path tracking in multi-axis High Speed Machining
Authors:
David Prévost,
Sylvain Lavernhe,
Claire Lartigue,
Didier Dumur
Abstract:
Within the context of High Speed Machining, it is essential to manage the trajectory generation to achieve both high surface quality and high productivity. As feed drives are one part of the set Machine tool - Numerical Controller, it is necessary to improve their performances to optimize feed drive dynamics during trajectory follow up. Hence, this paper deals with the modelling of the feed drive…
▽ More
Within the context of High Speed Machining, it is essential to manage the trajectory generation to achieve both high surface quality and high productivity. As feed drives are one part of the set Machine tool - Numerical Controller, it is necessary to improve their performances to optimize feed drive dynamics during trajectory follow up. Hence, this paper deals with the modelling of the feed drive in the case of multi axis machining. This model can be used for the simulation of axis dynamics and tool-path tracking to tune parameters and optimize new frameworks of command strategies. A procedure of identification based on modern NC capabilities is presented and applied to industrial HSM centres. Efficiency of this modelling is assessed by experimental verifications on various representative trajectories. After implementing a Generalized Predictive Control, reliable simulations are performed thanks to the model. These simulations can then be used to tune parameters of this new framework according to the tool-path geometry.
△ Less
Submitted 16 July, 2011;
originally announced July 2011.
