The Vision Transformer has attained remarkable success in various computer vision applications. However, the large computational costs and complex design limit its ability in handling large feature maps. Existing research predominantly focuses on constraining attention to small local regions, which reduces the number of tokens attending the attention computation while overlooking computational demands caused by the feed-forward layer in the Vision Transformer block. In this paper, we introduce Group Vision Transformer (GVT), a relatively simple and efficient variant of Vision Transformer, aiming to improve attention computation. The core idea of our model is to divide and group the entire Transformer layer, instead of only the attention part, into multiple independent branches. This approach offers two advantages: (1) It helps reduce parameters and computational complexity; (2) it enhances the diversity of the learned features. We conduct comprehensive analysis of the impact of different numbers of groups on model performance, as well as their influence on parameters and computational complexity. Our proposed GVT demonstrates competitive performances in several common vision tasks. For example, our GVT-Tiny model achieves 84.8% top-1 accuracy on ImageNet-1K, 51.4% box mAP and 45.2% mask mAP on MS COCO object detection and instance segmentation, and 50.1% mIoU on ADE20K semantic segmentation, outperforming the CAFormer-S36 model by 0.3% in ImageNet-1K top-1 accuracy, 1.2% in box mAP, 1.0% in mask mAP on MS COCO object detection and instance segmentation, and 1.2% in mIoU on ADE20K semantic segmentation, with similar model parameters and computational complexity. Code is accessible at https://github.com/yaoppeng/GVT.

Modern deep neural networks have achieved great successes in medical image analysis. However, the features captured by convolutional neural networks (CNNs) or Transformers tend to be optimized for pixel intensities and neglect key anatomical structures such as connected components and loops. In this paper, we propose a persistent homology guided approach (PHG-Net) that explores topological features of objects for medical image classification. For an input image, we first compute its cubical persistence diagram and extract topological features into a vector representation using a small neural network (called the PH module). The extracted topological features are then incorporated into the feature map generated by CNN or Transformer for feature fusion. The PH module is lightweight and capable of integrating topological features into any CNN or Transformer architectures in an end-to-end fashion. We evaluate our PHG-Net on three public datasets and demonstrate its considerable improvements on the target classification tasks over state-of-the-art methods.

Understanding of spatial attributes is central to effective 3D radiology image analysis where crop-based learning is the de facto standard. Given an image patch, its core spatial properties (e.g., position & orientation) provide helpful priors on expected object sizes, appearances, and structures through inherent anatomical consistencies. Spatial correspondences, in particular, can effectively gauge semantic similarities between inter-image regions, while their approximate extraction requires no annotations or overbearing computational costs. However, recent 3D contrastive learning approaches either neglect correspondences or fail to maximally capitalize on them. To this end, we propose an extensible 3D contrastive framework (Spade, for Spa tial De biasing) that leverages extracted correspondences to select more effective positive & negative samples for representation learning. Our method learns both globally invariant and locally equivariant representations with downstream segmentation in mind. We also propose separate selection strategies for global & local scopes that tailor to their respective representational requirements. Compared to recent state-of-the-art approaches, Spade shows notable improvements on three downstream segmentation tasks (CT Abdominal Organ, CT Heart, MR Heart).

Visual relationship is crucial to image understanding and can be applied to many tasks (e.g., image caption and visual question answering). Despite great progress on many vision tasks, relationship detection remains a challenging problem due to the complexity of modeling the widely spread and imbalanced distribution of {subject - predicate - object} triplets. In this paper, we propose a new framework to capture the relative positions and sizes of the subject and object in the feature map and add a new branch to filter out some object pairs that are unlikely to have relationships. In addition, an activation function is trained to increase the probability of some feature maps given an object pair. Experiments on two large datasets, the Visual Relationship Detection (VRD) and Visual Genome (VG) datasets, demonstrate the superiority of our new approach over state-of-the-art methods. Further, ablation study verifies the effectiveness of our techniques.