refineNet_multi-path refinement network for high-resolution semantic seg
0. 写作目的 好记性不如烂笔头。 1. 问题 下采样导致损失了细节信息。 2. 解决方法 高层的信息和低层的信息都有用。 作者对所有的信息都使用。 3. 模型 具体细节： 2x表示使用两次 RCU。逐层进行使用refineNet，如图2中的C图所示，对于下采样32倍的feature，refineNet只有一个输入。对于其他的refineNet则有两个输入。 Chained Res
相关文章
- 1. GCN_large kernel matters_improving semantic seg by global conv network
- 2. 语义分割--Label Refinement Network for Coarse-to-Fine Semantic Segmentation
- 3. RefineNet: Multi-Path Refinement Networks for High-Resolution Semantic Segmentation
- 4. Single-Shot Refinement Neural Network for Object Detection
- 5. Single-Shot Refinement Neural Network for Object Detection[reading notes]
- 6. PBRNet:Progressive Boundary Refinement Network for Temporal Action Detection （AAAI 2020）
- 7. Face paper：Single-Shot Refinement Neural Network for Object Detection[CVPR2018]
- 8. Reading Note: Single-Shot Refinement Neural Network for Object Detection
- 9. Pyramid Attention Network for Semantic Segmentation
- 10. Learning Deconvolution Network for Semantic Segmentation
- 更多相关文章...
- • Swift for 循环 - Swift 教程
- • Scala for循环 - Scala教程
- • Composer 安装与使用
- • Flink 数据传输及反压详解
相关标签/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
欢迎关注本站公众号,获取更多信息
相关文章
- 1. GCN_large kernel matters_improving semantic seg by global conv network
- 2. 语义分割--Label Refinement Network for Coarse-to-Fine Semantic Segmentation
- 3. RefineNet: Multi-Path Refinement Networks for High-Resolution Semantic Segmentation
- 4. Single-Shot Refinement Neural Network for Object Detection
- 5. Single-Shot Refinement Neural Network for Object Detection[reading notes]
- 6. PBRNet:Progressive Boundary Refinement Network for Temporal Action Detection （AAAI 2020）
- 7. Face paper：Single-Shot Refinement Neural Network for Object Detection[CVPR2018]
- 8. Reading Note: Single-Shot Refinement Neural Network for Object Detection
- 9. Pyramid Attention Network for Semantic Segmentation
- 10. Learning Deconvolution Network for Semantic Segmentation