2017/11/17 by Weiyue Wang, Qiangui Huang, Wang, Weiyue +7 · 24 citations
Computer Science · Engineering · #3D Shape Modeling and Analysis #Advanced Vision and Imaging #Adversarial system #Artificial intelligence #Artificial neural network #Computer Graphics and Visualization Techniques #Computer science #Convolutional neural network #Deep learning #Encoder #Generative adversarial network #Generative grammar #Image (mathematics) #Inpainting #Network architecture #Pattern recognition (psychology) #Recurrent neural network #Sequence (biology) #cs.CV
paper · pdf · doi:10.48550/arxiv.1711.06375
published in arXiv (Cornell University) (Cornell University)
arxiv created 2017/11/17 · openalex publication_date 2017/11/17 · arxiv updated 2017/11/20 · openalex created_date 2017/12/04 · openalex updated_date 2026/08/05
Recent advances in convolutional neural networks have shown promising results in 3D shape completion. But due to GPU memory limitations, these methods can only produce low-resolution outputs. To inpaint 3D models with semantic plausibility and contextual details, we introduce a hybrid framework that combines a 3D Encoder-Decoder Generative Adversarial Network (3D-ED-GAN) and a Long-term Recurrent Convolutional Network (LRCN). The 3D-ED-GAN is a 3D convolutional neural network trained with a generative adversarial paradigm to fill missing 3D data in low-resolution. LRCN adopts a recurrent neural network architecture to minimize GPU memory usage and incorporates an Encoder-Decoder pair into a Long Short-term Memory Network. By handling the 3D model as a sequence of 2D slices, LRCN transforms a coarse 3D shape into a more complete and higher resolution volume. While 3D-ED-GAN captures global contextual structure of the 3D shape, LRCN localizes the fine-grained details. Experimental results on both real-world and synthetic data show reconstructions from corrupted models result in complete and high-resolution 3D objects.