DL之DeconvNet：DeconvNet算法的簡介(論文介紹)、架構詳解、案例應用等配圖集合之詳細攻略

?

?

?

?

目錄

DeconvNet算法的簡介(論文介紹)

0、實驗結果

DeconvNet算法的架構詳解

DeconvNet算法的案例應用

---

?

?

?

?

?

?

?

相關文章
DL之DeconvNet：DeconvNet算法的簡介(論文介紹)、架構詳解、案例應用等配圖集合之詳細攻略
DL之DeconvNet：DeconvNet算法的架構詳解

?

DeconvNet算法的簡介(論文介紹)

? ? ? ? DeconvNet網絡架構，是由Convolution network、Deconvolution network兩種架構組成。

• Convolution network：feature extractor，采用VGG-16提取特征；
• Deconvolution network：shape generator，通過上采樣，計算像素的類別得分圖。
  ?

Abstract ?
? ? ? ?We propose a novel semantic segmentation algorithm by ?learning a deconvolution network. We learn the network ?on top of the convolutional layers adopted from VGG 16- ?layer net. The deconvolution network is composed of deconvolution ?and unpooling layers, which identify pixel-wise ?class labels and predict segmentation masks. We apply the ?trained network to each proposal in an input image, and ?construct the final semantic segmentation map by combining ?the results from all proposals in a simple manner. The ?proposed algorithm mitigates the limitations of the existing ?methods based on fully convolutional networks by integrating ?deep deconvolution network and proposal-wise ?prediction; our segmentation method typically identifies detailed ?structures and handles objects in multiple scales naturally. ?Our network demonstrates outstanding performance ?in PASCAL VOC 2012 dataset, and we achieve the best accuracy ?(72.5%) among the methods trained with no external ?data through ensemble with the fully convolutional network.
? ? ? ?本文提出了一種新的基于反卷積網絡的語義分割算法。我們學習了VGG 16層網在卷積層之上的網絡。反卷積網絡由反褶積層和反池層組成，它們識別像素級標簽并預測分割掩碼。我們將訓練好的網絡應用于輸入圖像中的每個提案，并將所有提案的結果以一種簡單的方式結合起來，構造出最終的語義分割圖。該算法將深度反卷積網絡與建議預測相結合，克服了現有全卷積網絡方法的局限性;我們的分割方法通常識別詳細的結構和處理對象在多個尺度自然。我們的網絡在PASCAL VOC 2012數據集中表現出色，通過全卷積網絡集成，在沒有外部數據訓練的方法中，我們的準確率最高(72.5%)。
Conclusion ?
? ? ? ?We proposed a novel semantic segmentation algorithm ?by learning a deconvolution network. The proposed deconvolution ?network is suitable to generate dense and pre-cise object segmentation masks since coarse-to-fine structures ?of an object is reconstructed progressively through ?a sequence of deconvolution operations. Our algorithm ?based on instance-wise prediction is advantageous to handle ?object scale variations by eliminating the limitation ?of fixed-size receptive field in the fully convolutional network. ?We further proposed an ensemble approach, which ?combines the outputs of the proposed algorithm and FCNbased ?method, and achieved substantially better performance ?thanks to complementary characteristics of both algorithms. ?Our network demonstrated the state-of-the-art ?performance in PASCAL VOC 2012 segmentation benchmark ?among the methods trained with no external data.
? ? ? ?本文提出了一種新的基于反卷積網絡的語義分割算法。該反褶積網絡通過一系列的反卷積操作，逐步重構出由粗到細的目標結構，適用于生成密集的預分割掩碼。我們的基于實例預測的算法消除了全卷積網絡中固定大小接受域的限制，有利于處理對象尺度變化。我們進一步提出了一種集成方法，將所提算法的輸出與基于FCN的方法相結合，由于兩種算法的互補特性，取得了較好的性能。在沒有外部數據訓練的方法中，我們的網絡在PASCAL VOC 2012分割基準測試中展示了最先進的性能。

?

論文
Hyeonwoo Noh, SeunghoonHong, BohyungHan.
Learning deconvolution network for semantic segmentation, ICLR, 2015.
https://arxiv.org/abs/1505.04366

?

0、實驗結果

1、PASCAL VOC 2012驗證圖像的語義分割結果實例
Example of semantic segmentation results on PASCAL VOC 2012 validation images
PASCAL VOC 2012上獲得的mean IoU=72.5%

(a) Examples that our method produces better results than FCN	? ? ? GT框是人工標定框、FCN算法、DeconvNet算法、EDeconvNet算法(FCN和EDeconvNet集成學習后)、EDeconvNet+CRF后處理的算法(效果更好！)
(b) Examples that FCN produces better results than our method 所提出的方法和FCN具有用于語義分割的互補特性，并且兩種方法的組合通過集成學習提高了準確性。	? ? ? ?下圖展示了FCN算法優于DeconvNet算法(因為多出了一些細節)，同樣的，EDeconvNet+CRF后處理的算法(效果更好！)
集成學習 (c) Examples that inaccurate predictions from our method and FCN are improved by ensemble	? ?

?

?

DeconvNet算法的架構詳解

更新……

DL之DeconvNet：DeconvNet算法的架構詳解

?

?

?

?

?

DeconvNet算法的案例應用

更新……

?

?

?

?

?

?

總結

以上是生活随笔為你收集整理的DL之DeconvNet：DeconvNet算法的简介(论文介绍)、架构详解、案例应用等配图集合之详细攻略的全部內容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網站內容還不錯，歡迎將生活随笔推薦給好友。