In this paper, we study the problem of 3D reconstruction from a single-view
RGB image and propose a novel approach called DIG3D for 3D object
reconstruction and novel view synthesis. Our method utilizes an encoder-decoder
framework which generates 3D Gaussians in decoder with the guidance of
depth-aware image features from encoder. In particular, we introduce the use of
deformable transformer, allowing efficient and effective decoding through 3D
reference point and multi-layer refinement adaptations. By harnessing the
benefits of 3D Gaussians, our approach offers an efficient and accurate
solution for 3D reconstruction from single-view images. We evaluate our method
on the ShapeNet SRN dataset, getting PSNR of 24.21 and 24.98 in car and chair
dataset, respectively. The result outperforming the recent method by around
2.25%, demonstrating the effectiveness of our method in achieving superior
results.

利用单视角 RGB 图像进行三维重建及新视角合成的研究中，提出了一种名为 DIG3D 的新方法，该方法利用编码 - 解码框架，通过编码器的深度感知图像特征指导解码器生成三维高斯图像，具体来说，引入了可变形变换器，通过三维参考点和多层细化适应来进行高效且有效的解码。通过利用三维高斯图像的优势，我们的方法为从单视角图像进行三维重建提供了高效且准确的解决方案。我们在 ShapeNet SRN 数据集上对方法进行了评估，在汽车和椅子数据集上获得了分别为 24.21 和 24.98 的峰值信噪比（PSNR），结果比最近的方法提高了约 2.25％，证明了我们方法在实现优越结果方面的有效性。