公开(公告)号：US20180005070A1

公开(公告)日：2018-01-04

申请号：US15705151

申请日：2017-09-14

Applicant: Adobe Systems Incorporated

Inventor： Zhe Lin ,  Xiaohui Shen ,  Jonathan Brandt ,  Jianming Zhang

IPC: G06K9/46 ,  G06K9/62 ,  G06K9/48 ,  G06N3/08 ,  G06N3/04

CPC classification number: G06K9/4628 ,  G06K9/6298 ,  G06K2009/00322 ,  G06N3/04 ,  G06N3/08

Abstract: Techniques for increasing robustness of a convolutional neural network based on training that uses multiple datasets and multiple tasks are described. For example, a computer system trains the convolutional neural network across multiple datasets and multiple tasks. The convolutional neural network is configured for learning features from images and accordingly generating feature vectors. By using multiple datasets and multiple tasks, the robustness of the convolutional neural network is increased. A feature vector of an image is used to apply an image-related operation to the image. For example, the image is classified, indexed, or objects in the image are tagged based on the feature vector. Because the robustness is increased, the accuracy of the generating feature vectors is also increased. Hence, the overall quality of an image service is enhanced, where the image service relies on the image-related operation.

公开(公告)号：US09846840B1

公开(公告)日：2017-12-19

申请号：US15164310

申请日：2016-05-25

Applicant: Adobe Systems Incorporated

Inventor： Zhe Lin ,  Xiaohui Shen ,  Jonathan W. Brandt ,  Jianming Zhang

IPC: G06N3/08 ,  G06F17/30

CPC classification number: G06N3/084 ,  G06F17/30259

Abstract: Semantic class localization techniques and systems are described. In one or more implementation, a technique is employed to back communicate relevancies of aggregations back through layers of a neural network. Through use of these relevancies, activation relevancy maps are created that describe relevancy of portions of the image to the classification of the image as corresponding to a semantic class. In this way, the semantic class is localized to portions of the image. This may be performed through communication of positive and not negative relevancies, use of contrastive attention maps to different between semantic classes and even within a same semantic class through use of a self-contrastive technique.

公开(公告)号：US09830526B1

公开(公告)日：2017-11-28

申请号：US15166164

申请日：2016-05-26

Applicant: Adobe Systems Incorporated

Inventor： Zhe Lin ,  Xiaohui Shen ,  Jonathan Brandt ,  Jianming Zhang

IPC: G06K9/62 ,  G06K9/46 ,  G06K9/48 ,  G06N3/08 ,  G06N3/04

CPC classification number: G06K9/4628 ,  G06K9/6298 ,  G06K2009/00322 ,  G06N3/04 ,  G06N3/08

Abstract: Techniques for increasing robustness of a convolutional neural network based on training that uses multiple datasets and multiple tasks are described. For example, a computer system trains the convolutional neural network across multiple datasets and multiple tasks. The convolutional neural network is configured for learning features from images and accordingly generating feature vectors. By using multiple datasets and multiple tasks, the robustness of the convolutional neural network is increased. A feature vector of an image is used to apply an image-related operation to the image. For example, the image is classified, indexed, or objects in the image are tagged based on the feature vector. Because the robustness is increased, the accuracy of the generating feature vectors is also increased. Hence, the overall quality of an image service is enhanced, where the image service relies on the image-related operation.

公开(公告)号：US09805445B2

公开(公告)日：2017-10-31

申请号：US14524489

申请日：2014-10-27

Applicant: Adobe Systems Incorporated

Inventor： Zhe Lin ,  Radomir Mech ,  Xiaohui Shen ,  Brian L. Price ,  Jianming Zhang

IPC: G06T3/40

CPC classification number: G06T3/40

Abstract: Image zooming is described. In one or more implementations, zoomed croppings of an image are scored. The scores calculated for the zoomed croppings are indicative of a zoomed cropping's inclusion of content that is captured in the image. For example, the scores are indicative of a degree to which a zoomed cropping includes salient content of the image, a degree to which the salient content included in the zoomed cropping is centered in the image, and a degree to which the zoomed cropping preserves specified regions-to-keep and excludes specified regions-to-remove. Based on the scores, at least one zoomed cropping may be chosen to effectuate a zooming of the image. Accordingly, the image may be zoomed according to the zoomed cropping such that an amount the image is zoomed corresponds to a scale of the zoomed cropping.

公开(公告)号：US09760978B1

公开(公告)日：2017-09-12

申请号：US15149415

申请日：2016-05-09

Applicant: Adobe Systems Incorporated

Inventor： Xin Lu ,  Zhe Lin ,  Jen-Chan Jeff Chien

IPC: G06T5/00 ,  G06T5/50 ,  G01R33/56 ,  G06T9/00 ,  H04N1/60

CPC classification number: G06T5/001 ,  G01R33/56 ,  G06T5/005 ,  G06T5/006 ,  G06T5/50 ,  G06T9/002 ,  G06T2207/10016 ,  G06T2207/20081 ,  G06T2207/20084 ,  G06T2207/20212 ,  H04N1/60

Abstract: Missing region prediction techniques are described. In implementations, an image pair is obtained that includes first and second images. The first image is corrupted by removing a region of content, resulting in a corrupted image having a missing region. The corrupted image and the second image of the image pair are then used to generate a training-image pair. Then, based on a plurality of training-image pairs including the generated training-image pair, a model is trained using machine learning. The model can subsequently be used to predict pixel values of pixels within a subsequent missing region of a subsequent image that is not used as part of the training.

公开(公告)号：US20170139572A1

公开(公告)日：2017-05-18

申请号：US14944019

申请日：2015-11-17

Applicant: Adobe Systems Incorporated

Inventor： Kalyan K. Sunkavalli ,  Zhe Lin ,  Xiaohui Shen ,  Joon-Young Lee

IPC: G06F3/0484 ,  G06T11/00 ,  G06F3/0488

CPC classification number: G06F3/04845 ,  G06F3/04842 ,  G06F3/0488 ,  G06T11/001

Abstract: In embodiments of image color and tone style transfer, a computing device implements an image style transfer algorithm to generate a modified image from an input image based on a color style and a tone style of a style image. A user can select the input image that includes color features, as well as select the style image that includes an example of the color style and the tone style to transfer to the input image. A chrominance transfer function can then be applied to transfer the color style to the input image, utilizing a covariance of an input image color of the input image to control modification of the input image color. A luminance transfer function can also be applied to transfer the tone style to the input image, utilizing a tone mapping curve based on a non-linear optimization to estimate luminance parameters of the tone mapping curve.

公开(公告)号：US09563825B2

公开(公告)日：2017-02-07

申请号：US14549350

申请日：2014-11-20

Applicant: Adobe Systems Incorporated

Inventor： Xiaohui Shen ,  Haoxiang Li ,  Zhe Lin ,  Jonathan W. Brandt

IPC: G06K9/66 ,  G06N3/04 ,  G06K9/46 ,  G06K9/62

CPC classification number: G06K9/66 ,  G06K9/4628 ,  G06K9/6267 ,  G06K9/6272 ,  G06N3/04 ,  G06N3/0454

Abstract: A convolutional neural network is trained to analyze input data in various different manners. The convolutional neural network includes multiple layers, one of which is a convolution layer that performs a convolution, for each of one or more filters in the convolution layer, of the filter over the input data. The convolution includes generation of an inner product based on the filter and the input data. Both the filter of the convolution layer and the input data are binarized, allowing the inner product to be computed using particular operations that are typically faster than multiplication of floating point values. The possible results for the convolution layer can optionally be pre-computed and stored in a look-up table. Thus, during operation of the convolutional neural network, rather than performing the convolution on the input data, the pre-computed result can be obtained from the look-up table.

Abstract translation: 训练卷积神经网络以各种不同的方式分析输入数据。 卷积神经网络包括多个层，其中之一是卷积层，其对于卷积层中的一个或多个滤波器的每一个，通过输入数据执行滤波器的卷积。 卷积包括基于滤波器和输入数据生成内积。 卷积层的滤波器和输入数据都被二值化，允许使用通常快于浮点值乘法的特定运算来计算内积。 可以可选地预先计算卷积层的可能结果并将其存储在查找表中。 因此，在卷积神经网络的操作期间，不是对输入数据执行卷积，可以从查找表中获得预先计算的结果。

公开(公告)号：US20160379091A1

公开(公告)日：2016-12-29

申请号：US14747877

申请日：2015-06-23

Applicant: Adobe Systems Incorporated

Inventor： Zhe Lin ,  Stefan Guggisberg ,  Jonathan Brandt ,  Michael Marth

IPC: G06K9/62 ,  G06K9/52 ,  G06K9/66

CPC classification number: G06K9/6256 ,  G06K9/00724 ,  G06K9/6262

Abstract: This disclosure relates to training a classifier algorithm that can be used for automatically selecting tags to be applied to a received image. For example, a computing device can group training images together based on the training images having similar tags. The computing device trains a classifier algorithm to identify the training images as semantically similar to one another based on the training images being grouped together. The trained classifier algorithm is used to determine that an input image is semantically similar to an example tagged image. A tag is generated for the input image using tag content from the example tagged image based on determining that the input image is semantically similar to the tagged image.

Abstract translation: 本公开涉及训练可用于自动选择要应用于接收到的图像的标签的分类器算法。 例如，计算设备可以基于具有相似标签的训练图像将训练图像组合在一起。 计算设备训练分类器算法，以基于训练图像被分组在一起来将训练图像识别为语义上彼此相似。 训练分类器算法用于确定输入图像在语义上类似于示例标记图像。 基于确定输入图像在语义上类似于带标签的图像，使用来自示例标记图像的标签内容为输入图像生成标签。

公开(公告)号：US20160292537A1

公开(公告)日：2016-10-06

申请号：US15183629

申请日：2016-06-15

Applicant: Adobe Systems Incorporated

Inventor： Xiaohui Shen ,  Zhe Lin ,  Jonathan W. Brandt

IPC: G06K9/62 ,  G06N99/00 ,  G06N5/04 ,  G06K9/46 ,  G06F17/30

CPC classification number: G06K9/6256 ,  G06F17/30247 ,  G06F17/30253 ,  G06K9/46 ,  G06K9/4676 ,  G06K9/469 ,  G06N5/04 ,  G06N99/005

Abstract: Feature interpolation techniques are described. In a training stage, features are extracted from a collection of training images and quantized into visual words. Spatial configurations of the visual words in the training images are determined and stored in a spatial configuration database. In an object detection stage, a portion of features of an image are extracted from the image and quantized into visual words. Then, a remaining portion of the features of the image are interpolated using the visual words and the spatial configurations of visual words stored in the spatial configuration database.

Abstract translation: 描述特征插值技术。 在训练阶段，从训练图像的集合中提取特征并量化为视觉词。 训练图像中的视觉词的空间配置被确定并存储在空间配置数据库中。 在物体检测阶段，从图像中提取图像的特征的一部分并量化为视觉词。 然后，使用存储在空间配置数据库中的视觉词和视觉词的空间配置来内插图像的特征的剩余部分。

公开(公告)号：US09436893B2

公开(公告)日：2016-09-06

申请号：US14091972

申请日：2013-11-27

Applicant: Adobe Systems Incorporated

Inventor： Jianchao Yang ,  Zhaowen Wang ,  Zhe Lin ,  Jonathan Brandt

IPC: G06K9/62

CPC classification number: G06K9/6269 ,  G06K9/6235

Abstract: A system and method for distributed similarity learning for high-dimensional image features are described. A set of data features is accessed. Subspaces from a space formed by the set of data features are determined using a set of projection matrices. Each subspace has a dimension lower than a dimension of the set of data features. Similarity functions are computed for the subspaces. Each similarity function is based on the dimension of the corresponding subspace. A linear combination of the similarity functions is performed to determine a similarity function for the set of data features.

Abstract translation: 描述了用于高维图像特征的分布式相似性学习的系统和方法。 访问一组数据功能。 使用一组投影矩阵来确定由该组数据特征形成的空间的子空间。 每个子空间的尺寸小于数据特征集合的维度。 为子空间计算相似度函数。 每个相似度函数基于相应子空间的维度。 执行相似度函数的线性组合以确定该组数据特征的相似度函数。