Topic
Geometric feature learning
About: Geometric feature learning is a research topic. Over the lifetime, 5 publications have been published within this topic receiving 1 citations.
Papers
20 Jun 2021
TL;DR: Li et al. as discussed by the authors proposed a two-stream graph convolutional network (TSGCNet) to learn multi-view geometric information from different geometric attributes for 3D shape segmentation.
Abstract: The ability to segment teeth precisely from digitized 3D dental models is an essential task in computer-aided orthodontic surgical planning. To date, deep learning based methods have been popularly used to handle this task. State-of-the-art methods directly concatenate the raw attributes of 3D inputs, namely coordinates and normal vectors of mesh cells, to train a single-stream network for fully-automated tooth segmentation. This, however, has the drawback of ignoring the different geometric meanings provided by those raw attributes. This issue might possibly confuse the network in learning discriminative geometric features and result in many isolated false predictions on the dental model. Against this issue, we propose a two-stream graph convolutional network (TSGCNet) to learn multi-view geometric information from different geometric attributes. Our TSGCNet adopts two graph-learning streams, designed in an input-aware fashion, to extract more discriminative high-level geometric representations from coordinates and normal vectors, respectively. These feature representations learned from the designed two different streams are further fused to integrate the multi-view complementary information for the cell-wise dense prediction task. We evaluate our proposed TSGCNet on a real-patient dataset of dental models acquired by 3D intraoral scanners, and experimental results demonstrate that our method significantly outperforms state-of-the-art methods for 3D shape segmentation.
45 citations
1 Jun 2021
TL;DR: In this paper, a batch-neighborhood normalization (BNN) is proposed to improve robustness to mean-std variation of local feature distributions that presumably can happen in samples with varying point density.
Abstract: Extracting geometric features from 3D models is a common first step in applications such as 3D registration, tracking, and scene flow estimation. Many hand-crafted and learning-based methods aim to produce consistent and distinguishable geometric features for 3D models with partial overlap. These methods work well in cases where the point density and scale of the overlapping 3D objects are similar, but struggle in applications where 3D data are obtained independently with unknown global scale and scene overlap. Unfortunately, instances of this resolution mismatch are common in practice, e.g., when aligning data from multiple sensors. In this work, we introduce a new normalization technique, Batch-Neighborhood Normalization, aiming to improve robustness to mean-std variation of local feature distributions that presumably can happen in samples with varying point density. We empirically demonstrate that the presented normalization method’s performance compares favorably to comparison methods in indoor and outdoor environments, and on a clinical dataset, on common point registration benchmarks in both standard and, particularly, resolution-mismatch settings. The source code and clinical dataset are available at https://github.com/lppllppl920/NeighborhoodNormalization-Pytorch.
6 citations
TL;DR: A bio-inspired method that utilizes the line-segment-based representation to perform a dedicated channel for the geometric feature learning process is proposed and is found to be effective for classifying the hand-drawn sketches.
2 citations
Posted Content•
TL;DR: Li et al. as discussed by the authors proposed a two-stream graph convolutional network (TSGCNet) to learn multi-view geometric information from different geometric attributes for 3D shape segmentation.
Abstract: The ability to segment teeth precisely from digitized 3D dental models is an essential task in computer-aided orthodontic surgical planning. To date, deep learning based methods have been popularly used to handle this task. State-of-the-art methods directly concatenate the raw attributes of 3D inputs, namely coordinates and normal vectors of mesh cells, to train a single-stream network for fully-automated tooth segmentation. This, however, has the drawback of ignoring the different geometric meanings provided by those raw attributes. This issue might possibly confuse the network in learning discriminative geometric features and result in many isolated false predictions on the dental model. Against this issue, we propose a two-stream graph convolutional network (TSGCNet) to learn multi-view geometric information from different geometric attributes. Our TSGCNet adopts two graph-learning streams, designed in an input-aware fashion, to extract more discriminative high-level geometric representations from coordinates and normal vectors, respectively. These feature representations learned from the designed two different streams are further fused to integrate the multi-view complementary information for the cell-wise dense prediction task. We evaluate our proposed TSGCNet on a real-patient dataset of dental models acquired by 3D intraoral scanners, and experimental results demonstrate that our method significantly outperforms state-of-the-art methods for 3D shape segmentation.
1 Aug 2021
TL;DR: In this article, the dominant facial regions of interest are extracted and used to further learn spatio-temporal features with space-time auto-correlation of gradients (STACOG) technique.
Abstract: When an emotional state is involuntarily and spontaneously delivered with low intensity and short duration, micro expression (ME) occurs. Developing from psychological perspectives to computer vision standpoints, ME has obtained huge advancements and breakthroughs. A variety of feature extraction techniques have been introduced based on different outlooks. With exclusive characteristics of ME, geometric feature learning is involved in approaching the problem in this paper. Specifically, we propose a method where dominant facial regions of interest are extracted and used to further learn spatio-temporal features with space-time auto-correlation of gradients (STACOG) technique. The facial motion features are fed into a multi-layer perceptron network for emotion classification. The combination of ROIs and STACOG captures ME with respect to geometrical property along side with the temporal aspect in three dimensional space. This puts a stress on presumably suboptimal features, making them more salient and resilient for the classification stage. The framework is experimented on three well-known, state-of-the-art spontaneous ME databases CASMEII, SMIC and SAM.
Performance Metrics
| Year | Papers |
|---|---|
| 2021 | 3 |
| 2020 | 2 |