Yue Zhong's research while affiliated with Beijing University of Posts and Telecommunications and other places

Publications (6)

Article
Deep image-based modeling has received a lot of attention in recent years. Sketch-based modeling in particular has gained popularity given the ubiquitous nature of touchscreen devices. In this paper, we (i) study and compare diverse single-image reconstruction methods on sketch input, comparing the different 3D shape representations: multi-view, vo...
Article
In this paper, for the first time, we investigate the problem of generating 3D shapes from professional 2D sketches via deep learning. We target sketches done by professional artists, as these sketches are likely to contain more details than the ones produced by novices, and thus the reconstruction from such sketches poses a higher demand on the le...
Preprint
Full-text available
Deep image-based modeling received lots of attention in recent years, yet the parallel problem of sketch-based modeling has only been briefly studied, often as a potential application. In this work, for the first time, we identify the main differences between sketch and image inputs: (i) style variance, (ii) imprecise perspective, and (iii) sparsit...

Citations

... ShapeMVD [Lun et al. 2017] ✗ ✗ ✗ Pixel2Mesh ✓ ✗ ✗ ProSketch [Zhong et al. 2021] ✗ ✗ ✗ Sketch2Mesh [Guillard et al. 2021] ✓ ✓ ✗ DeepSketch [Zhong et al. 2022] ✓ ✗ ✗ Ours ✓ ✓ ✓ Table 1. Comparison between sketch-based shape generation methods. ...
... Thanks to the recent progress in deep learning, the understanding of freehand sketches and the quality of single-view generation have reached an unprecedented level. Several intelligent sketchbased modeling systems have been developed to enable novice users to create visually plausible 3D models within a few minutes [7], [8], [9]. Closest to our work, DeepSketch2Face [10] presents the first deep learning-based sketching system for modeling 3D faces by mapping 2D sketches into a parametric space for face generation. ...
... Besides, object shape mask is used in 3D sketch shape reconstruction. Zhong et al. [41] feed two sketches from different views into a generator that outputs a set of normal and depth maps and shape masks, where the shape masks are used to constrain the final 3D shape. Nie et al. [42] propose a feature fuse scheme that uses quantified mask features as attention weights to further optimize the shape recognition. ...
... Sketch can be used for human-computer interaction, thus sketch recognition has been the important research topic in recent years. Most of previous works focus on single-object sketch classification or retrieval [23,2,24,25,26], leaving scene sketch classification under-studied. In particular, scene sketch zero-shot learning (SSZSL) has not been studied to date. ...