专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

111. 发明专利

GB2572029A Detecting objects using a weakly supervised model 未知
公开(公告)号：GB2572029A
公开(公告)日：2019-09-18
申请号：GB201817588
申请日：2018-10-29
申请人： ADOBE INC
发明人： DELUN DU , ZHE LIN , BALDO FAIETA
IPC分类号： G06K9/00 , G06K9/46 , G06N3/02
摘要： The invention provides systems and methods for detecting an object in an input image based on a target object keyword (e.g. people). A heat map 310a of the input image 306 is generated based on the target object keyword. Various bounding boxes 314a, based on a pixel analysis of the heat map, are then derived. One or more centre boxes 316a are then generated from the bounding boxes using local thresholds. Object location proposals 318a,b,c are also determined based on the keyword. The bounding boxes are then used to determine scores for the object location proposals and the box 318c for the highest scoring object location proposal is overlaid on the input image. The object location proposals could be derived using a YOLO algorithm. The heat map could be generated using a fully-convolutional dense tagging network. The method could also employ a weakly supervised model which avoids some of the costs of a learning process.

112. 发明专利

GB2571825A Semantic class localization digital environment 未知
公开(公告)号：GB2571825A
公开(公告)日：2019-09-11
申请号：GB201900596
申请日：2019-01-16
申请人： ADOBE INC
发明人： ZHE LIN , YUFEI WANG , XIAOHUI SHEN , SCOTT D COHEN , JIANMING ZHANG
IPC分类号： G06K9/62 , G06K9/00 , G06T7/10
摘要： Computer-implemented segmentation techniques are disclosed for identifying features in an image based on text tags. The method comprises: converting a tag 118 into a vector representation 202, the tag defining a semantic class 120 to be located in a digital image 106; generating an attention map 210 by an embedding neural network 208 based on the digital image and the vector representation, the attention map defining a location in the digital image that corresponds to the semantic class, where the embedding neural network has been trained using image-level tags of respective semantic classes; refining 212 the location of the semantic class in the attention map by a refinement neural network 214 , the refinement neural network trained using localized tags of respective semantic classes; and indicating the refined location of the semantic class in the digital image using the refined attention map 216. The method thus makes use of tags for similar items in a machine learning environment in order to provide more examples for training.

113. 发明专利

AU2018271418A1 CREATING SELECTIVE VIRTUAL LONG-EXPOSURE IMAGES 未知
公开(公告)号：AU2018271418A1
公开(公告)日：2019-08-29
申请号：AU2018271418
申请日：2018-12-02
申请人： ADOBE INC
发明人： SAFDARNEJAD SEYED MORTEZA , KONG SARAH AYE , WILENSKY GREGG DARRYL , HSIEH CHIH-YAO
IPC分类号： H04N1/00 , H04N7/00
摘要： Embodiments of the present invention are directed to facilitating images with selective application of the long-exposure effect. In accordance with some embodiments of the present invention, virtual long-exposure image comprising a plurality of aligned frames is provided and a selection of a region of pixels in the virtual long-exposure image is received. The virtual long-exposure image is combined with one of the frames forming the virtual long-exposure image to create a selective virtual long-exposure image. The selective virtual long-exposure image comprises a visible portion of the original virtual long-exposure image and a visible portion of the individual frame that corresponds to the selected region of pixels. Additional frames may be combined with the virtual long exposure image to create a plurality of selective virtual long-exposure image options, and the user may select one for continued use or for saving. U) C)LU 5; > DuJ 0 (D z z 0) LU LU 0/ 0 U : (N LUJ - c rHO LU

114. 发明专利

GB2571398A Refining local parameterizations for applying two-dimensional images to three-dimensional models 未知
公开(公告)号：GB2571398A
公开(公告)日：2019-08-28
申请号：GB201817899
申请日：2018-11-01
申请人： ADOBE INC
发明人： EMILIANO GAMBARETTO , QINGNAN ZHOU , VLADIMIR KIM , MEHMET YUMER
IPC分类号： G06T15/04
摘要： A method suitable for refining local parameterizations that apply two-dimensional images 212 to three-dimensional models 210 comprises: selecting, from a set of parameterization-initialization processes, a parameterization-initialization process based on feature(s) of a target mesh region 214, 216, 218 of the 3D model 210; generating an initial local parameterization, comprising a mapping of image points from a 2D image 212 to model points along a surface of the target mesh region, from the parameterization-initialization process; computing a quality metric of the initial local parameterization; generating a modified local parameterization by changing the mapping from the initial local parameterization; computing quality metric of the modified local parameterization; and modifying the 3D model by applying the image points to the model points in accordance with the modified local parameterization. The set of parameterization-initialization processes may comprise an exponential map parameterization, harmonic map parameterization, and prior parameterization for a different mesh region of the 3D model that neighbours or overlaps the target mesh region. The quality metric may indicate an amount of distortion in the 2D image as applied to the target mesh region. Generating the modified local parameterization may comprise performing an iterative refinement of the mapping from the initial local parameterization.

115. 发明专利

GB201909236D0 Utilizing a dynamic memory network to track digital dialog states and generate responses 未知
公开(公告)号：GB201909236D0
公开(公告)日：2019-08-14
申请号：GB201909236
申请日：2019-06-27
申请人： ADOBE INC
摘要： The present disclosure relates to generating digital responses based on digital dialog states generated by a neural network having a dynamic memory network architecture. For example, in one or more embodiments, the disclosed system provides a digital dialog having one or more segments to a dialog state tracking neural network having a dynamic memory network architecture that includes a set of multiple memory slots. In some embodiments, the dialog state tracking neural network further includes update gates and reset gates used in modifying the values stored in the memory slots. For instance, the disclosed system can utilize cross-slot interaction update/reset gates to accurately generate a digital dialog state for each of the segments of digital dialog. Subsequently, the system generates a digital response for each segment of digital dialog based on the digital dialog state.

116. 发明专利

GB2570751A Predicting style breaches within textual content 未知
公开(公告)号：GB2570751A
公开(公告)日：2019-08-07
申请号：GB201814860
申请日：2018-09-13
申请人： ADOBE INC
发明人： BASAVA RAJ K , VIVEK GUPTA , PRANAV R MANERIKER , ANANDHAVELU NATARAJAN
IPC分类号： G06F17/21 , G06F16/30 , G06F16/95 , G06F17/27
摘要： Method for facilitating predicting style breaches within content. The method involves obtaining target content comprising content created by a plurality of individuals, identifying style features associated with a plurality of content segments within the target content and then using the style features and a style breach prediction model to predict style breaches within the content. The style breaches indicating a change the writing style of text within the target content. The style features may comprise lexical features and/or syntactic features. The lexical features may comprise one or more of an average word length, a sentence length, a word length frequency, a Flesh-Kincaid readability score, a frequency of words not in the English dictionary, a Honore’s Index value, a Hapax legomena value, a Hapax dislegomena value, a Yule’s Index value and a token ratio. Also disclosed is a method for facilitating a style breach prediction model by collecting training content, identifying style features in the training content, obtaining style breach annotations associated with the training content, the annotations indicating positions at which style within the training content is perceived as different and training a style breach prediction model to predict changes in text style within content using this method.

117. 发明专利

AU2018202767A1 DATA STRUCTURE AND ALGORITHM FOR TAG LESS SEARCH AND SVG RETRIEVAL 未知
公开(公告)号：AU2018202767A1
公开(公告)日：2019-07-11
申请号：AU2018202767
申请日：2018-04-20
申请人： ADOBE INC
发明人： BISWAS SANJEEV KUMAR
IPC分类号： G06T7/00
摘要： Approaches are described for shape-based graphics search. Each graphics object of a set of graphics objects is analyzed. The analyzing includes determining an outline of the graphics object from graphics data that forms the graphics object. The outline of the graphics object is sampled resulting in sampled points that capture the outline of the graphics object. A shape descriptor of the graphics object is determined which captures local and global geometric properties of the sampled points. Search results of a search query are determined based on a comparison between a shape descriptor of a user identified graphics object and the shape descriptor of at least one graphics object of the set of graphics objects. At least one of the search results can be presented on a user device associated with the search query.

118. 发明专利

DE102018007941A1 Generieren eines Dreiecksnetzes für ein Bild, das durch Kurven dargestellt wird 未知
公开(公告)号：DE102018007941A1
公开(公告)日：2019-07-04
申请号：DE102018007941
申请日：2018-10-08
申请人： ADOBE INC
发明人： BATRA VINEET , FISHER MATTHEW DAVID , WAMPLER KEVIN JOHN , KAUFMAN DANIEL M , PHOGAT ANKIT
IPC分类号： G06T11/60 , G06T17/20
摘要： Beschrieben werden hier Systeme und Techniken zum Generieren eines Dreiecksnetzes für ein Bild, das durch Kurven (beispielsweise Bezier-Segmente) dargestellt wird. Ein Umriss eines Bildes wird bestimmt und auf einen Satz von verbundenen Linienzügen verringert, die effizient in einer Kantenliste dargestellt werden. Ein Dreiecksnetz wird auf Grundlage der Kantenliste anstatt durch direktes Abtasten der Kurven des Bildes und Nutzen der Abtastungen als Vertices von Dreiecken generiert. Das Dreiecksnetz wird mit einer Anzahl von Dreiecken generiert, die von einer Anzahl von Kurven, die das Bild darstellen, unabhängig ist. Abtastungen der Kurven werden mit dem Dreiecksnetz gekoppelt, indem die Abtastungen mit baryzentrischen Koordinaten in Bezug auf ein Dreieck in dem Netz dargestellt werden. Sobald ein Netz verformt wird, werden Orte der Abtastungen aus den baryzentrischen Koordinaten und Dreiecken in dem verformten Netz bestimmt und zum Rekonstruieren der Kurven des verformten Bildes genutzt.

119. 发明专利

GB2569858A Constructing content based on multi-sentence compression of source content 未知
公开(公告)号：GB2569858A
公开(公告)日：2019-07-03
申请号：GB201817487
申请日：2018-10-26
申请人： ADOBE INC
发明人： NATWAR MODANI , KUNDAN KRISHNA , PRANAV RAVINDRA MANERIKER , BALAJI VASAN SRINIVASAN
IPC分类号： G06F16/31
摘要： Embodiments of the present disclosure provide systems, methods, and computer storage media directed to facilitating corpus-based content generation, in particular, using graph-based multi-sentence compression to generate a final content output. The method involves obtaining source content relevant to an input text snippet (e.g. keywords, phrases or sentences), utilizing a graphical representation of a plurality of sentences form the source content to identify a set of sentences in the snippet having overlapping information, generating a candidate sentence by compressing content in the set of sentences having overlapping information and then generating and providing final content comprising a set of candidate sentences including the candidate sentence. The method may also involve parsing the source content relevant to the input snippet into a plurality of sentences. Each of the plurality of sentences may be mapped to a first graph and assigned a weight, wherein the weight relates to the relevance between the corresponding sentence and the input snippet. At least a portion of the weighted sentences may be parsed into a plurality of word tokens.

120. 发明专利

GB2569847A Removing and replacing objects in images according to a directed user conversation 未知
公开(公告)号：GB2569847A
公开(公告)日：2019-07-03
申请号：GB201817165
申请日：2018-10-22
申请人： ADOBE INC
发明人： SCOTT D COHEN , BRIAN LYNN PRICE , ABHINAV GUPTA
IPC分类号： G06T5/00 , G06F40/00 , G06T11/60 , G06T19/20 , G10L15/08
摘要： Systems and techniques are described herein for directing a user conversation to obtain an editing query, and removing and replacing objects in an image based on the editing query. Pixels corresponding to an object in the image indicated by the editing query are ascertained. The editing query is processed to determine whether it includes a remove request or a replace request. A search query is constructed to obtain images, such as from a database of stock images, including fill material or replacement material to fulfill the remove request or replace request, respectively. Composite images are generated from the fill material or the replacement material and the image to be edited. Composite images arc harmonized to remove editing artifacts and make the images look natural. A user interface exposes images, and the user interface accepts multi-modal user input during the directed user conversation.

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