Towards Efficient Local 3D Conditioning
SessionPosters Gallery
DescriptionWe propose an innovative weight-encoded, locally conditioned neural implicit representation, utilizing a neural network to approximate a grid of latent codes, while sharing the decoder across the entire category. This approach significantly enhances reconstruction quality compared to global methods while minimizing memory usage for storing geometries.
Event Type
Poster
TimeThursday, 14 December 202310:00am - 5:30pm
LocationExhibition Hall 1, Level 2 (Exhibition Centre)
