[經典網絡]：Mask R-CNN 網絡結構詳解 : R-CNN, Fast R-CNN, Faster R-CNN, Mask R-CNN, FCN

論文：He, Kaiming, et al. “Mask r-cnn.” Proceedings of the IEEE international conference on computer vision. 2017.

代碼：TensorFlow實作

執行個體分割模型Mask R-CNN詳解：從R-CNN，Fast R-CNN，Faster R-CNN再到Mask R-CNN從R-CNN1開講講到Mask R-CNN2，很完整，簡明的表述了R-CNN系列的發展，但是不夠細節。Mask R-CNN2可以了解為Faster R-CNN3 + FCN4

1. R-CNN2算法詳解

a. 在圖像中确定約1000-2000個候選框

b. 對于每個候選框内圖像塊，使用深度網絡提取特征

c. 對候選框中提取出的特征，使用分類器判别是否屬于一個特定類

d. 對于屬于某一特征的候選框，用回歸器進一步調整其位置

1. Fast R-CNN5算法詳解

網絡末端同步訓練的分類和位置調整，提升準确度

使用多尺度的圖像金字塔，性能幾乎沒有提高

倍增訓練資料，能夠有2%-3%的準确度提升

網絡直接輸出各類機率(softmax)，比SVM分類器性能略好

更多候選窗不能提升性能

1. Faster R-CNN3算法詳解

目标檢測的四個基本步驟（候選區域生成，特征提取，分類，位置精修）終于被統一到一個深度網絡架構之内

1. Mask R-CNN算法詳解

   

   輸入圖檔進行特征提取，使用FPN6網絡

[5]:Lin, Tsung-Yi, et al. “Feature pyramid networks for object detection.” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2017.

參考：

1. Girshick, Ross, et al. “Rich feature hierarchies for accurate object detection and semantic segmentation.” Proceedings of the IEEE conference on computer vision and pattern recognition. 2014. ↩︎
2. He, Kaiming, et al. “Mask r-cnn.” Proceedings of the IEEE international conference on computer vision. 2017. ↩︎ ↩︎ ↩︎
3. Ren, Shaoqing, et al. “Faster r-cnn: Towards real-time object detection with region proposal networks.” Advances in neural information processing systems. 2015. ↩︎ ↩︎
4. Long, Jonathan, Evan Shelhamer, and Trevor Darrell. “Fully convolutional networks for semantic segmentation.” Proceedings of the IEEE conference on computer vision and pattern recognition. 2015. ↩︎
5. Girshick, Ross. “Fast r-cnn.” Proceedings of the IEEE international conference on computer vision. 2015. ↩︎
6. Lin, Tsung-Yi, et al. “Feature pyramid networks for object detection.” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2017. ↩︎