Intracerebral hemorrhage (ICH) is a life-threatening stroke that requires precise segmentation for effective treatment. To aid in diagnosing ICH, various deep learning-based methods have been proposed. However, these methods face challenges posed by irregular patterns and low-contrast boundaries in ICH images. In this paper, we introduce a novel conditional diffusion-based segmentation approach for ICH segmentation. Our framework leverages ResNet18 and a Transformer Block to enhance conditional feature extraction, as well as cross-attention mechanisms to align global and local features between the conditional and diffusion branches. Experiment results on the Instance2022 dataset show an improvement exceeding 10% in the Dice coefficient and IoU score, compared to the baseline diffusion-based model, MedSegDiffv2. More impressively, more than 20% improvement is achieved in terms of both the Dice coefficient and IoU score in cross-dataset evaluations on two benchmarks, i.e., the BHSD and PhysioNet, demonstrating excellent generalizability to unseen data.

Point cloud shape completion, which aims to reconstruct the missing regions of the incomplete point clouds with plausible shapes, is an ill-posed and challenging task that benefits many downstream 3D applications. Prior approaches achieve this goal by employing a two-stage completion framework, generating a coarse yet complete seed point cloud through an encoder-decoder network, followed by refinement and upsampling. However, the encoded features suffer from information loss of the missing portion, leading to an inability of the decoder to reconstruct seed points with detailed geometric clues. To tackle this issue, we propose a novel Orthogonal Dictionary Guided Shape Completion Network (ODGNet). The proposed ODGNet consists of a Seed Generation U-Net, which leverages multi-level feature extraction and concatenation to significantly enhance the representation capability of seed points, and Orthogonal Dictionaries that can learn shape priors from training samples and thus compensate for the information loss of the missing portions during inference. Our design is simple but to the point, extensive experiment results indicate that the proposed method can reconstruct point clouds with more details and outperform previous state-of-the-art counterparts. The implementation code is available at https://github.com/corecai163/ODGNet.

Cross-modal 2D-3D point cloud semantic segmentation using few-shot-based learning provides a practical approach for borrowing matured 2D domain knowledge into the 3D segmentation model, which reduces the reliance on laborious 3D annotation work and improves generalization to new categories. However, previous methods use single-view point cloud generation algorithms to bridge the gap between 2D images and 3D point clouds, leaving the incomplete geometry of an object or scene due to occlusions. To address this issue, we propose a novel view synthesis cross-modal few-shot point cloud semantic segmentation network. It introduces the color and depth inpainting to generate multi-view images and masks, which compensate for the absent depth information of generated point clouds. Additionally, we propose a Co-embedding Network to bridge the domain features between synthesized and original, collected 3D data, and a weighted prototype network is employed to balance the impact of multi-view images and enhance the segmentation performance. Extensive experiments on two benchmarks show the superiority of our method by outperforming the existing cross-modal few-shot 3D segmentation methods.

Abstract Based on the coding mechanism and interactive features of visual information in the visual pathway, a new method of image contour detection is proposed. Firstly, simulating the visual adaptation characteristics of retinal ganglion cells, an adaptation & sensitization regulation model (ASR) based on the adaptation-sensitization characteristics is proposed, which introduces a sinusoidal function curve modulated by amplitude, frequency and initial phase to dynamically adjusted color channel response information and enhance the response of color edges. Secondly, the color antagonism characteristic is introduced to process the color edge responses, and the obtained primary contour responses is fed forward to the dorsal pathway across regions. Then, the coding characteristics of the “angle” information in the V2 region are simulated, and a double receptive fields model (DRFM) is constructed to compensate for the missing detailed contours in the generation of primary contour responses. Finally, a new double stream information fusion model (DSIF) is proposed, which simulates the dorsal overall contour information flow by the across-region response weighted fusion mechanism, and introduces the multi-directional fretting to simulate the fine-tuning characteristics of ventral detail features simultaneously, extracting the significant contours by weighted fusion of dorsal and ventral information streams. In this paper, the natural images in BSDS500 and NYUD datasets are used as experimental data, and the average optimal F-score of the proposed method is 0.72 and 0.69, respectively. The results show that the proposed method has better results in texture suppression and significant contour extraction than the comparison method.

Designing a point cloud upsampler, which aims to generate a clean and dense point cloud given a sparse point representation, is a fundamental and challenging problem in computer vision. A line of attempts achieves this goal by establishing a point-to-point mapping function via deep neural networks. However, these approaches are prone to produce outlier points due to the lack of explicit surface-level constraints. To solve this problem, we introduce a novel surface regularizer into the upsampler network by forcing the neural network to learn the underlying parametric surface represented by bicubic functions and rotation functions, where the new generated points are then constrained on the underlying surface. These designs are integrated into two different networks for two tasks that take advantages of upsampling layers -- point cloud upsampling and point cloud completion for evaluation. The state-of-the-art experimental results on both tasks demonstrate the effectiveness of the proposed method. The implementation code will be available at https://github.com/corecai163/PSCU.