Aerial semantic segmentation suffers from the class imbalance problem. Proper differentiation of least represented categories is challenging and a goal for the state-of-art approach. In this work, we present a novel deep learning method to perform this task. We proposed a lightweight encoder-...
First, we introduce a deep learning based algorithm for multi-class semantic segmentation addressing fourteen different tissue types in whole-slide images of colorectal cancer, including not only the primary cancer-associated epithelial and stroma classes but also some other more peripheral tissue types....
这篇文章提出了一个新的基于transformer的框架,为 weakly supervised semantic segmenta-tion (WSSS) 任务学习一个作为伪标签的 class-specific object localization maps 。受 standard vision transformer 中单类标记的参与区域可以用来形成与类无关的定位图这一事实的启发,作者研究了 transformer 模型是否也能通过学习tr...
在以往的研究中,通常将VIT中一个class token和所有的patch token一起参与训练,在训练结束后,单独提取class token做下游任务(分类、分割等),但是在多标签任务(multi-label)和语义分割(Semantic Segmentation)等任务中,用单个class token来实现常常效果不佳,因为单个class token只能指导与类无关(class-agnostic)的定位映...
1. Weakly Supervised Semantic Segmentation 概念 弱监督语义分割(Weakly Supervised Semantic Segmentation, WSSS)是一种图像处理技术,旨在使用较弱的标签(如图像级标签)来训练模型,以实现图像中每个像素的类别预测。与传统的全监督语义分割相比,弱监督语义分割显著降低了标注成本,但在性能上可能会有所下降。
Multi-class Token Transformer for Weakly Supervised Semantic Segmentation -Supplementary Material- A. Implementation details A.1. Training and testing of MCTformer To integrate the CAM module into the proposed MCT- former, we used a convolutional layer with C kernels of 3 × 3, a stride of...
Segmentation-based multi-class semantic object detection. Multimedia Tools and Applications, 60(2):305-326, 2012.Vieux, R., Benois-Pineau, J., Domenger, J.P., Braquelaire, A.: Segmentation- based multi-class semantic object detection. Multimedia Tools and Ap- plications pp. 1-22-22 (...
As a data-driven method, the performance of deep convolutional neural networks (CNN) relies heavily on training data. The prediction results of traditional networks give a bias toward larger classes, which tend to be the background in the semantic segmentation task. This becomes a major problem ...
I am using the tiramisu architecture for semantic segmentation which uses negative log likelihood as the loss (implementation here: https://github.com/bfortuner/pytorch_tiramisu). The results so far are great. I highly recommend using this architecture for semantic segmentation. Have not tried it ...
As a data-driven method, the performance of deep convolutional neural networks (CNN) relies heavily on training data. The prediction results of traditional networks give a bias toward larger classes, which tend to be the background in the semantic segmentation task. This becomes a major problem ...