Shape Registration in the Time of Transformers
Dec 6, 2021
Speakers
Giovanni Trappolini
Speaker · 0 followers
Luca Cosmo
Speaker · 0 followers
Luca Moschella
Speaker · 0 followers
Riccardo Marin
Speaker · 0 followers
Simone Melzi
Speaker · 0 followers
Emanuele Rodola
Speaker · 0 followers
About
In this paper, we propose a transformer-based procedure for the efficient registration of non-rigid 3D point clouds. The proposed approach is data-driven and adopts for the first time the transformers architecture in the registration task. Our method is general and applies to different settings. Given a fixed template with some desired properties (e.g. skinning weights or other animation cues), we can register raw acquired data to it, thereby transferring all the template properties to the input geometry. Alternatively, given a pair of shapes, our method can register the first onto the second (or vice-versa), obtaining a high-quality dense correspondence between the two.In both contexts, the quality of our results enables us to target real applications such as texture transfer and shape interpolation.Furthermore, we also show that including an estimation of the underlying density of the surface eases the learning process. By exploiting the potential of this architecture, we can train our model requiring only a sparse set of ground truth correspondences (10∼20% of the total points). The proposed model and the analysis that we perform pave the way for future exploration of transformer-based architectures for registration and matching applications. Qualitative and quantitative evaluations demonstrate that our pipeline outperforms state-of-the-art methods for deformable and unordered 3D data registration on different datasets and scenarios.In this paper, we propose a transformer-based procedure for the efficient registration of non-rigid 3D point clouds. The proposed approach is data-driven and adopts for the first time the transformers architecture in the registration task. Our method is general and applies to different settings. Given a fixed template with some desired properties (e.g. skinning weights or other animation cues), we can register raw acquired data to it, thereby transferring all the template properties to the input…
Organizer
NeurIPS 2021
Account · 1.9k followers
About NeurIPS 2021
Neural Information Processing Systems (NeurIPS) is a multi-track machine learning and computational neuroscience conference that includes invited talks, demonstrations, symposia and oral and poster presentations of refereed papers. Following the conference, there are workshops which provide a less formal setting.
Like the format? Trust SlidesLive to capture your next event!
Professional recording and live streaming, delivered globally.
Sharing
Recommended Videos
Presentations on similar topic, category or speaker
Data preparation for training CNNs: Application to vibration-based condition monitoring
Watch later
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Understanding, Predicting, and Manipulating Image Memorability with Representation Learning
Watch later
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Understanding the Effects of Dataset Characteristics on Offline Reinforcement Learning
Watch later
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
G-PATE: Scalable Differentially Private Data Generator via Private Aggregation of Teacher Discriminators
Watch later
Yunhui Long, …
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
STORM+: Fully Adaptive SGD with Momentum for Nonconvex Optimization
Watch later
Kfir Y. Levy, …
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Language models enable zero-shot prediction of the effects of mutations on protein function
Watch later
Joshua Meier, …
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%