Existing zero-shot 3D point cloud segmentation methods often struggle with limited transferability from seen classes to unseen classes and from semantic to visual space. To alleviate this, we introduce 3D-PointZshotS, a geometry-aware zero-shot segmentation framework that enhances both feature generation and alignment using latent geometric prototypes (LGPs). Specifically, we integrate LGPs into a generator via a cross-attention mechanism, enriching semantic features with fine-grained geometric details. To further enhance stability and generalization, we introduce a self-consistency loss, which enforces feature robustness against point-wise perturbations. Additionally, we re-represent visual and semantic features in a shared space, bridging the semantic-visual gap and facilitating knowledge transfer to unseen classes. Experiments on three real-world datasets, namely ScanNet, SemanticKITTI, and S3DIS, demonstrate that our method achieves superior performance over four baselines in terms of harmonic mIoU. Code is available at github.

Existing few-shot image classification networks aim to perform prediction on images belonging to classes that were not seen during training, with only a few labeled images, which are randomly picked from the same image pool as the support set. However, this traditional approach has two main issues: (i) in real-world applications, since support images are randomly picked, the angle they were captured from can be very different from that of the query image, causing the images to look very different and making it hard to match them; (ii) since support and query images, for both training and testing, are sampled from the same image pool, models can overfit the dataset, especially if the image pool contains images with similar color, texture or view angle. Thus, good performance on a dataset does not reflect a model’s real ability. To address these issues, we propose a novel few-shot learning approach referred to as the 3D guided 2D (3DG2D) few-shot image classification. In our proposed approach, the queries are 2D images, and the support set is composed of 3D mesh data, providing different views of an object, in contrast to randomly picked images providing a single view. From each 3D mesh, 14 projection images are generated from different angles. Thus, these projections have significant variance among themselves. To address this challenge, we also propose the Angle Inference Module (AIM), which is used to infer the view angle of a query image so that more attention is given to projection images corresponding to the same view angle as the query image to achieve better prediction performance. We perform experiments on ModelNet40, Toys4K and ShapeNet datasets with 4-fold cross validation, and show that our 3DG2D few-shot classification approach consistently outperforms the state-of-the-art baselines.

Few-shot learning often assumes that base classes are abundant and diverse with plentiful well-labeled samples for each class. This ensures that models can generalize effectively from a small amount of data by leveraging prior knowledge learned from base classes. This assumption holds for 2D few-shot learning since the benchmark datasets are large and diverse. However, 3D point cloud few-shot benchmarks are low in magnitude and diversity. We conduct experiments and show that many existing methods overlook this issue and suffer from overfitting on base classes, which hinders generalization ability and test performance. To alleviate the overfitting issue, we propose a simplified manifold mixup, referred to as the SimpliMix, which mixes hidden representations and forces the models to learn more generalized features. We incorporate SimpliMix into existing prototype-based models, perform experiments on ModelNet40-FS, ModelNet40-C-FS and ScanObjectNN-FS datasets, and improve the models by a significant margin. We further conduct cross-domain few-shot classification experiments and show that networks with SimpliMix learn more generalized and transferable features and achieve better performance. The code is available at https://github.com/LexieYang/SimpliMix

Although different approaches have been proposed for 3D point cloud-related tasks, few-shot learning (FSL) of 3D point clouds still remains under-explored. In FSL, un-like traditional supervised learning, the classes of training and test data do not overlap, and a model needs to rec-ognize unseen classes from only a few samples. Existing FSL methods for 3D point clouds employ point-based models as their backbone. Yet, based on our extensive experiments and analysis, we first show that using a point-based backbone is not the most suitable FSL approach, since (i) a large number of points' features are discarded by the max pooling operation used in 3D point-based backbones, decreasing the ability of representing shape information; (ii) point-based backbones are sensitive to occlusion. To address these issues, we propose employing a projection-and 2D Convolutional Neural Network-based backbone, referred to as the ViewNet, for FSL from 3D point clouds. Our approach first projects a 3D point cloud onto six different views to alleviate the issue of missing points. Also, to generate more descriptive and distinguishing features, we propose View Pooling, which combines different projected plane combinations into five groups and performs max-pooling on each of them. The experiments performed on the ModelNet40, ScanObjectNN and ModelNet40-C datasets, with cross validation, show that our method consistently outperforms the state-of-the-art baselines. Moreover, compared to traditional image classification backbones, such as ResNet, the proposed ViewNet can extract more distinguishing features from multiple views of a point cloud. We also show that ViewNet can be used as a backbone with different FSL heads and provides improved performance compared to traditionally used backbones.

Recent years have witnessed significant progress in the field of few-shot image classification while few-shot 3D point cloud classification still remains under-explored. Real-world 3D point cloud data often suffers from occlusions, noise and deformation, which make the few-shot 3D point cloud classification even more challenging. In this paper, we propose a cross-modality feature fusion network, for few-shot 3D point cloud classification, which aims to recognize an object given only a few labeled samples, and provides better performance even with point cloud data with missing points. More specifically, we train two models in parallel. One is a projection-based model with ResNet18 as the backbone and the other one is a point-based model with a DGCNN backbone. Moreover, we design a Support-Query Mutual Attention (sqMA) module to fully exploit the correlation between support and query features. Extensive experiments on three datasets, namely ModelNet40, ModelNet40-C and ScanObjectNN, show the effectiveness of our method, and its robustness to missing points. Our proposed method outperforms different state-of-the-art baselines on all datasets. The margin of improvement is even larger on the ScanObjectNN dataset, which is collected from real-world scenes and is more challenging with objects having missing points.

Haze removal is important to improve performance of outdoor vision systems. However, it is challenging to remove haze from a single nighttime haze image. In this paper, a novel superpixel-based single image haze removal algorithm is proposed for nighttime haze images. The input nighttime image is first decomposed into a glow image and a glow-free nighttime haze image using their relative smoothness. A superpixel-based method is then introduced to compute the value of the atmospheric light and dark channel for each pixel in the glow-free haze image. The transmission map is decomposed from the dark channel of the glow-free haze image by the weighted guided image filter. Since superpixels usually adhere to the boundaries of objects well, a smaller local window size can be selected. As such, details in areas of fine structures are preserved better. In addition, to avoid noticeable noise in the sky area, an adaptive threshold is added to the transmission map when the nighttime haze image is restored. Experiments show that our method produces better results than the existing haze removal algorithms for nighttime haze images.