专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

101. 发明专利

GB2521958B System and method for producing edited images using embedded plug-in 未知
公开(公告)号：GB2521958B
公开(公告)日：2019-12-04
申请号：GB201506224
申请日：2013-09-13
申请人： ADOBE INC
发明人： ABRAHAM MOSHE MUCHNICK , ISRAEL DERDIK
IPC分类号： G06T11/60 , G06F3/0484

102. 发明专利

GB2574087A Compositing aware digital image search 未知
公开(公告)号：GB2574087A
公开(公告)日：2019-11-27
申请号：GB201903414
申请日：2019-03-13
申请人： ADOBE INC
发明人： XIAOHUI SHEN , ZHE LIN , KALYAN KRISHNA SUNKAVALLI , HENGSHUANG ZHAO , BRIAN LYNN PRICE
IPC分类号： G06F16/58 , G06N3/08 , G06N20/00 , G06T7/10
摘要： A compositing aware digital image search system wherein a background feature machine learning system and a foreground feature machine learning system extracts background and foreground features from a digital image 402-404, a score is calculated through feature embedding based on the extracted features from a plurality of candidate digital images 408 and a search is performed to output a search result 410. The machine learning systems may be convolutional neural networks. There may be a category feature machine learning system to generate category features from categorical data which may be a vector representation of text. There is also a method wherein a positive foreground and background image is extracted from a single digital image (1102, Figure 11), a positive background digital image is generated by filling a region in the background digital image from which the foreground image is extracted (1104, Figure 11), obtaining a negative foreground image and training background and foreground feature machine learning systems jointly using a loss function based on the positive foreground image, the positive background image and the negative foreground image (1106-1110, Figure 11).

103. 发明专利

AU2019200721A1 ONLINE TRAINING AND UPDATE OF FACTORIZATION MACHINES USING ALTERNATING LEAST SQUARES OPTIMIZATION 未知
公开(公告)号：AU2019200721A1
公开(公告)日：2019-11-14
申请号：AU2019200721
申请日：2019-02-02
申请人： ADOBE INC
发明人： MITRA SAAYAN , MAO XUEYU , SWAMINATHAN VISWANATHAN , SARKHEL SOMDEB , LI SHENG
IPC分类号： G06N20/10
摘要： Techniques are disclosed for training of factorization machines (FMs) using a streaming mode alternating least squares (ALS) optimization. A methodology implementing the techniques according to an embodiment includes receiving a datapoint that includes a feature vector and an associated target value. The feature vector includes user identification, subject matter identification, and a context. The target value identifies an opinion of the user relative to the subject matter. The method further includes applying an FM to the feature vector to generate an estimate of the target value, and updating parameters of the FM for training of the FM. The parameter update is based on application of a streaming mode ALS optimization to: the datapoint; the estimate of the target value; and to an updated summation of intermediate calculated terms generated by application of the streaming mode ALS optimization to previously received datapoints associated with prior parameter updates of the FM. Online Streaming Data Streaming mode 110 ALS Optimization . Training and Update Module Datapoint 120 (x, y) Update Parameters '_ _Estimate Factorization A o Machine y Initial Parameters Batch mode ALS Optimization Training Module

104. 发明专利

GB2556412B Generating efficient, stylized mesh deformations using a plurality of input meshes 未知
公开(公告)号：GB2556412B
公开(公告)日：2019-10-23
申请号：GB201714998
申请日：2017-09-18
申请人： ADOBE INC
发明人： KEVIN WAMPLER
IPC分类号： G06T19/20 , G06T17/20

105. 发明专利

GB201912592D0 Environment map generation and hole filling 未知
公开(公告)号：GB201912592D0
公开(公告)日：2019-10-16
申请号：GB201912592
申请日：2019-09-02
申请人： ADOBE INC
摘要： In some embodiments, an image manipulation application receives a two-dimensional background image and projects the background image onto a sphere to generate a sphere image. Based on the sphere image, an unfilled environment map containing a hole area lacking image content can be generated. A portion of the unfilled environment map can be projected to an unfilled projection image using a map projection. The unfilled projection image contains the hole area. A hole filling model is applied to the unfilled projection image to generate a filled projection image containing image content for the hole area. A filled environment map can be generated by applying an inverse projection of the map projection on the filled projection image and by combining the unfilled environment map with the generated image content for the hole area of the environment map.

106. 发明专利

GB2546368B Modelling semantic concepts in an embedding space as distributions 未知
公开(公告)号：GB2546368B
公开(公告)日：2019-10-02
申请号：GB201619341
申请日：2016-11-15
申请人： ADOBE INC
发明人： HAILIN JIN , ZHOU REN , ZHE LIN , CHEN FANG
IPC分类号： G06F16/583

107. 发明专利

AU2019200270A1 CONCEPT MASK: LARGE-SCALE SEGMENTATION FROM SEMANTIC CONCEPTS 未知
公开(公告)号：AU2019200270A1
公开(公告)日：2019-09-26
申请号：AU2019200270
申请日：2019-01-16
申请人： ADOBE INC
发明人： LIN ZHE , SHEN XIAOHUI , ZHANG JIANMING , COHEN SCOTT DAVID , WANG YUFEI
IPC分类号： G06T7/10 , G06N3/02
摘要： Semantic segmentation techniques and systems are described that overcome the challenges of limited availability of training data to describe the potentially millions of tags that may be used to describe semantic classes in digital images. In one example, the techniques are configured to train neural networks to leverage different types of training datasets using sequential neural networks and use of vector representations to represent the different semantic classes. Computing Device 102 Image Processing System 104 Semantic Digital Image Segmentation Class 120System116 Indication 122 Attention Map 24- Network (1rk) Digital Image 106 ?74 1

108. 发明专利

GB201911506D0 Video inpainting via confidence-weighted motion estimation 未知
公开(公告)号：GB201911506D0
公开(公告)日：2019-09-25
申请号：GB201911506
申请日：2019-08-12
申请人： ADOBE INC
摘要： A method of accessing a video having a target region 306, updating the video content in the target region based on confidence-weighted motion estimation for the target region, and presenting the updated video content on a display device. In particular, the video has a scene comprising a first 112a and a second 112b frame, the scene having an annotation identifying a target region to be modified in one or more video frames. A boundary motion for a boundary of the target region in the scene is computed, the boundary including boundary pixels 404a, 405a neighbouring the target region. Confidence values are assigned to the boundary pixels, wherein a confidence value is based on: a difference between a forward and reverse motion with respect to a particular boundary pixel, and/or a texture in a region that includes the particular boundary pixel. A target motion 412 of a target pixel is interpolated from the boundary motion, wherein the confidence value of the boundary pixel controls a contribution of a motion of the boundary pixel to the target motion. The colour of the target pixel is updated to correspond to the interpolated target motion.

109. 发明专利

GB201911502D0 Space-time memory network for locating target object in video content 未知
公开(公告)号：GB201911502D0
公开(公告)日：2019-09-25
申请号：GB201911502
申请日：2019-08-12
申请人： ADOBE INC
摘要： A space-time memory (i.e. neural) network locates target object(s) in video content for segmentation or other object classification. Claimed is generating (204) a query key map and a query value map by applying a space-time memory network to a query frame (202) depicting a target feature and retrieving (206), from a memory, a memory key map and a memory value map that are computed from a set of memory frames from video content that includes the query frame. Memory weights are computed (208) by applying a similarity function to the memory key map and the query key map and the space-time memory network is used to classify (210) (e.g. generate a segmentation mask) content in the query frame as depicting the target feature based on a weighted summation that includes the memory weights applied to memory locations in the memory value map. Also claimed is accessing, from video content, a query frame having content depicting a target feature, and performing a step for classifying content of the query frame as depicting the target feature by applying a space-time memory (i.e. neural) network to the query frame and one or more memory frames.

110. 发明专利

DE102018010163A1 Automatisches Generieren sinnvoller Nutzersegmente 未知
公开(公告)号：DE102018010163A1
公开(公告)日：2019-09-19
申请号：DE102018010163
申请日：2018-12-28
申请人： ADOBE INC
发明人： MODARRESI KOUROSH , YUAN HONGYUAN , MENGUY CHARLES
IPC分类号： G06F17/00 , G06Q50/00 , G16Z99/00
摘要： Offenbart werden Systeme, Verfahren und nichttemporäre computerlesbare Medien (Systeme) zum Generieren sinnvoller und aufschlussreicher Nutzersegmentberichte auf Grundlage eines hochdimensionalen Datenraumes. Insbesondere nutzen die offenbarten Systeme bei einer oder mehreren Ausführungsformen ein Modell des relaxierten Biclusterns zum automatischen Identifizieren von Nutzersegmenten in einem Datenraum, der Datensätze von Merkmalen beinhaltet, die für einzelne Nutzer spezifisch sind. Die offenbarten Systeme identifizieren und beinhalten Nutzer in automatisch generierten Nutzersegmenten beispielsweise auch dann, wenn diese Nutzer einigen, jedoch gegebenenfalls nicht allen Merkmalen wie andere Mitglieder in den automatisch generierten Nutzersegmenten zugeordnet sind.

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式