G3DR: Generative 3D Reconstruction in ImageNet
This addresses the need for efficient 3D asset generation from images, with significant performance gains but likely incremental improvements over existing generative 3D reconstruction methods.
The paper tackles the problem of generating diverse and high-quality 3D objects from single images, introducing G3DR which improves over state-of-the-art methods by up to 22% in perceptual metrics and 90% in geometry scores while requiring only half the training time.
We introduce a novel 3D generative method, Generative 3D Reconstruction (G3DR) in ImageNet, capable of generating diverse and high-quality 3D objects from single images, addressing the limitations of existing methods. At the heart of our framework is a novel depth regularization technique that enables the generation of scenes with high-geometric fidelity. G3DR also leverages a pretrained language-vision model, such as CLIP, to enable reconstruction in novel views and improve the visual realism of generations. Additionally, G3DR designs a simple but effective sampling procedure to further improve the quality of generations. G3DR offers diverse and efficient 3D asset generation based on class or text conditioning. Despite its simplicity, G3DR is able to beat state-of-theart methods, improving over them by up to 22% in perceptual metrics and 90% in geometry scores, while needing only half of the training time. Code is available at https://github.com/preddy5/G3DR