![]() The new features of version 7 are explained in separate pages: Instant Text 7 Pro New Features. The main features of Instant Text are presented in the Overview and in the pages on Key Features. Extensive experiments on a wide variety of benchmark datasets demonstrate that the proposed algorithm performs favorably against the state-of-the-art methods both qualitatively and quantitatively, while achieving significantly better efficiency. Instant Text 7 Pro is licensed through subscriptions and can be ordered online at this link. Finally, to address the Janus (multi-head) problem in 3D generation, we propose an adaptive Perp-Neg algorithm that can dynamically adjust its concept negation scales according to the severity of the Janus problem during training, effectively reducing the multi-head effect. Furthermore, we propose a simple yet effective activation function, the scaled-sigmoid, to replace the original sigmoid function, which speeds up the training convergence by more than ten times. The core innovation of our Instant3D lies in our exploration of strategies to effectively inject text conditions into the network. ![]() We achieve this remarkable speed by devising a new network that directly constructs a 3D triplane from a text prompt. Once trained, Instant3D is able to create a 3D object for an unseen text prompt in less than one second with a single run of a feedforward network. In this paper, we propose a novel framework for fast text-to-3D generation, dubbed Instant3D. This heavy and repetitive training cost impedes their practical deployment. Specifically, these methods optimize a neural field from scratch for each text prompt, taking approximately one hour or more to generate one object. While several existing works have achieved impressive results for this task, they mainly rely on a time-consuming optimization paradigm. ![]() Download a PDF of the paper titled Instant3D: Instant Text-to-3D Generation, by Ming Li and 6 other authors Download PDF Abstract:Text-to-3D generation, which aims to synthesize vivid 3D objects from text prompts, has attracted much attention from the computer vision community. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |