Over 20 Vector Institute Research Papers Accepted at CVPR 2023
The Vector Institute for Artificial Intelligence announced that 24 research papers co-authored by its Faculty Members and Affiliates were accepted at the IEEE/CVF Computer Vision and Pattern Recognition (CVPR) 2023 conference, held in Vancouver. This significant representation highlights the institute's contributions to AI and computer vision. Notable contributors include Raquel Urtasun and Sanja Fidler, who each co-authored five papers focusing on self-driving vehicle sensors and 3D environment construction, respectively. Key innovations featured include RobustNeRF, developed with David Fleet, which improves 3D scene representation from 2D images by ignoring transient objects, and Sparsifiner, created by Graham Taylor, a more efficient Vision Transformer model that reduces computational costs while maintaining accuracy in object recognition. Additionally, research on high-resolution video synthesis using Latent Diffusion Models was presented. These advancements demonstrate progress in neural radiance fields, efficient deep learning algorithms, and generative models, reinforcing the Vector Institute's role in advancing artificial intelligence research on a global stage.
Wire timeline
Over 20 Vector Institute Research Papers Accepted at CVPR 2023
The Vector Institute for Artificial Intelligence announced that 24 research papers co-authored by its Faculty Members and Affiliates were accepted at the IEEE/CVF Computer Vision and Pattern Recognition (CVPR) 2023 conference, held in Vancouver. This significant representation highlights the institute's contributions to AI and computer vision. Notable contributors include Raquel Urtasun and Sanja Fidler, who each co-authored five papers focusing on self-driving vehicle sensors and 3D environment construction, respectively. Key innovations featured include RobustNeRF, developed with David Fleet, which improves 3D scene representation from 2D images by ignoring transient objects, and Sparsifiner, created by Graham Taylor, a more efficient Vision Transformer model that reduces computational costs while maintaining accuracy in object recognition. Additionally, research on high-resolution video synthesis using Latent Diffusion Models was presented. These advancements demonstrate progress in neural radiance fields, efficient deep learning algorithms, and generative models, reinforcing the Vector Institute's role in advancing artificial intelligence research on a global stage.
Vector Institute for Artificial Intelligence