Detection guided deconvolutional network for hierarchical feature learning

Liu, Jing; Liu, Bingyuan; Lu, Hanqing

doi:10.1016/j.patcog.2015.02.002

Cited by 13 publications

(5 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Instead, these invariances are acquired through learning. Although the features extracted by DCNNs are significantly more powerful than their hand-designed counterparts like SIFT and HOG 20 23 , they may have difficulties to tackle 3-D transformations.…”

mentioning

confidence: 99%

Deep Networks Can Resemble Human Feed-forward Vision in Invariant Object Recognition

Kheradpisheh¹,

Ghodrati²,

Ganjtabesh³

et al. 2016

Sci Rep

162

108

View full text Add to dashboard Cite

Deep convolutional neural networks (DCNNs) have attracted much attention recently, and have shown to be able to recognize thousands of object categories in natural image databases. Their architecture is somewhat similar to that of the human visual system: both use restricted receptive fields, and a hierarchy of layers which progressively extract more and more abstracted features. Yet it is unknown whether DCNNs match human performance at the task of view-invariant object recognition, whether they make similar errors and use similar representations for this task, and whether the answers depend on the magnitude of the viewpoint variations. To investigate these issues, we benchmarked eight state-of-the-art DCNNs, the HMAX model, and a baseline shallow model and compared their results to those of humans with backward masking. Unlike in all previous DCNN studies, we carefully controlled the magnitude of the viewpoint variations to demonstrate that shallow nets can outperform deep nets and humans when variations are weak. When facing larger variations, however, more layers were needed to match human performance and error distributions, and to have representations that are consistent with human behavior. A very deep net with 18 layers even outperformed humans at the highest variation level, using the most human-like representations.

show abstract

mentioning

confidence: 99%

Deep Networks Can Resemble Human Feed-forward Vision in Invariant Object Recognition

Kheradpisheh¹,

Ghodrati²,

Ganjtabesh³

et al. 2016

Sci Rep

162

108

View full text Add to dashboard Cite

show abstract

“…For image analysis and synthesis, a more robust image feature representation can be learned by constructing a hierarchical deconvolution network. 50 This hierarchical deconvolution network structure helps the network extract the intermediate and advanced feature representations from the image.…”

Section: Layer Feature Fusion Methods Based On a Convolution-deconvol...mentioning

confidence: 99%

“…The convolution–deconvolution feature fusion method uses

1 \times 1

convolutions to fuse the convolution–deconvolution features and transfers the high‐level semantic feature information to the low‐level semantic feature information, enhancing the high‐level semantic features of the crack image while retaining the detail features and effectively eliminating the interference from complex background texture and noise features. For image analysis and synthesis, a more robust image feature representation can be learned by constructing a hierarchical deconvolution network 50 . This hierarchical deconvolution network structure helps the network extract the intermediate and advanced feature representations from the image.…”

Section: Overview Of Proposed Methodsmentioning

confidence: 99%

Concrete crack segmentation based on convolution–deconvolution feature fusion with holistically nested networks

Hao

Liu

et al. 2022

Structural Contr & Hlth

View full text Add to dashboard Cite

“…30,31 Many large CNNs with performance that can be scaled depending on the size of training data, model complexity, and processing power have achieved meaningful improvements in the object segmentation of images. [32][33][34][35][36][37][38][39] A fully convolutional network (FCN) is a deep learning network for image segmentation originally proposed in 2015. 39 Leveraging the advantages of convolutional computation in feature organization and extraction, an FCN establishes a multilayer convolution structure and reasonable sets deconvolution layer to realize pixel-by pixel segmentation.…”

Section: Segmentation Model Based On Convolutional Neural Network Fomentioning

confidence: 99%

Segmentation model based on convolutional neural networks for extracting vegetation from Gaofen-2 images

Zhang

Liu

Yang

et al. 2018

J. Appl. Rem. Sens.

View full text Add to dashboard Cite

Convolutional neural network (CNN) models achieve state-of-the-art performance for natural image semantic segmentation. An approach for extracting vegetation from Gaofen-2 (GF-2) remote sensing imagery based on the CNN model is presented. We constructed a convolutional encoder neural networks (CENN) consisting of two layers. The first layer has two sets of convolutional kernels for extracting the features of farmland and woodland, respectively. The second layer consists of two encoders that use nonlinear functions to encode the learned features and map the encoding results to the corresponding category number. In the training stage, samples of farmland, woodland, and other lands are categorically used to train the CENN. After training is accomplished, the CENN would acquire enough ability to accurately extract farmland and woodland from GF-2 imagery. The CENN was trained on 36 GF-2 images and tested on three other GF-2 images. We compared the proposed method to a deep belief network, a fully convolutional network, and a DeepLab model using the same images. The experiments demonstrate that the proposed approach improves upon the accuracy of existing approaches. The average precision, recall, and kappa coefficient of the proposed approach were 0.91, 0.87, and 0.86, respectively. Thus, the proposed approach is proven to effectively extract vegetation from GF-2 imagery.

show abstract

Detection guided deconvolutional network for hierarchical feature learning

Cited by 13 publications

References 26 publications

Deep Networks Can Resemble Human Feed-forward Vision in Invariant Object Recognition

Deep Networks Can Resemble Human Feed-forward Vision in Invariant Object Recognition

Concrete crack segmentation based on convolution–deconvolution feature fusion with holistically nested networks

Segmentation model based on convolutional neural networks for extracting vegetation from Gaofen-2 images

Contact Info

Product

Resources

About