The popularity of online home design and floor plan customization has been steadily increasing. However, the manual conversion of floor plan images from books or paper materials into electronic resources can be a challenging task due to the vast amount of historical data available. By leveraging neural networks to identify and parse floor plans, the process of converting these images into electronic materials can be significantly streamlined. In this paper, we present a novel learning framework for automatically parsing floor plan images. Our key insight is that the room type text is very common and crucial in floor plan images as it identifies the important semantic information of the corresponding room. However, this clue is rarely considered in previous learning-based methods. In contrast, we propose the Row and Column network (RC-Net) for recognizing floor plan elements by integrating the text feature. Specifically, we add the text feature branch in the network to extract text features corresponding to the room type for the guidance of room type predictions. More importantly, we formulate the Row and Column constraint module (RC constraint module) to share and constrain features across the entire row and column of the feature maps to ensure that only one type is predicted in each room as much as possible, making the segmentation boundaries between different rooms more regular and cleaner. Extensive experiments on three benchmark datasets validate that our framework substantially outperforms other state-of-the-art approaches in terms of the metrics of FWIoU, mACC and mIoU.

We address the 3D shape assembly of multiple geometric pieces without overlaps, a scenario often encountered in 3D shape design, field archeology, and robotics. Existing methods depend on strong assumptions on the number of shape pieces and coherent geometry or semantics of shape pieces. Despite raising attention to 3D registration with complex or low overlapping patterns, few methods consider shape assembly with rare overlaps. To address this problem, we present a novel framework inspired by solving puzzles, named PuzzleNet, which conducts multi-task learning by leveraging both 3D alignment and boundary information. Specifically, we design an end-to-end neural network based on a point cloud transformer with two-way branches for estimating rigid transformation and predicting boundaries simultaneously. The framework is then naturally extended to reassemble multiple pieces into a full shape by using an iterative greedy approach based on the distance between each pair of candidate-matched pieces. To train and evaluate PuzzleNet, we construct two datasets, named DublinPuzzle and ModelPuzzle, based on a real-world urban scan dataset (DublinCity) and a synthetic CAD dataset (ModelNet40) respectively. Experiments demonstrate our effectiveness in solving 3D shape assembly for multiple pieces with arbitrary geometry and inconsistent semantics. Our method surpasses state-of-the-art algorithms by more than 10 times in rotation metrics and four times in translation metrics.

Specular highlight detection and removal is a fundamental problem in computer vision and image processing. In this paper, we present an efficient end-to-end deep learning model for automatically detecting and removing specular highlights in a single image. In particular, an encoder–decoder network is utilized to detect specular highlights, and then a novel Unet-Transformer network performs highlight removal; we append transformer modules instead of feature maps in the Unet architecture. We also introduce a highlight detection module as a mask to guide the removal task. Thus, these two networks can be jointly trained in an effective manner. Thanks to the hierarchical and global properties of the transformer mechanism, our framework is able to establish relationships between continuous self-attention layers, making it possible to directly model the mapping between the diffuse area and the specular highlight area, and reduce indeterminacy within areas containing strong specular highlight reflection. Experiments on public benchmark and real-world images demonstrate that our approach outperforms state-of-the-art methods for both highlight detection and removal tasks.

A discriminative local shape descriptor plays an important role in various applications. In this paper, we present a novel deep learning framework that derives discriminative local descriptors for deformable 3D shapes. We use local "geometry images" to encode the multi-scale local features of a point, via an intrinsic parameterization method based on geodesic polar coordinates. This new parameterization provides robust geometry images even for badly-shaped triangular meshes. Then a triplet network with shared architecture and parameters is used to perform deep metric learning; its aim is to distinguish between similar and dissimilar pairs of points. Additionally, a newly designed triplet loss function is minimized for improved, accurate training of the triplet network. To solve the dense correspondence problem, an efficient sampling approach is utilized to achieve a good compromise between training performance and descriptor quality. During testing, given a geometry image of a point of interest, our network outputs a discriminative local descriptor for it. Extensive testing of non-rigid dense shape matching on a variety of benchmarks demonstrates the superiority of the proposed descriptors over the state-of-the-art alternatives.