-
71.发明申请公开(公告)号：US20180137835A1
公开(公告)日：2018-05-17
申请号：US15380414
申请日：2016-12-15
发明人： Siddhant Jain , Renzil Leith DSouza
摘要： Overlay and screen recording techniques are described that enables separate recordings of a screen and one or more overlays that were displayed on the screen during recording. In one example, pixel values of an overlay are blended with pixel values of a screen to paint the overlay onto the screen in a transparent manner that is imperceptible to the human eye but allows for original screen pixel values to be recovered from areas of the screen that were visually occluded by the overlay. This enables a user to display recording controls and visual cues on their screen without having to worry about the overlay visually occluding any screen content during the recording. One or both of the separately recorded screen and overlay streams can then be output for playback to enable viewing of the individual streams without loss in quality or content of the individual streams.
-
公开(公告)号：US09972111B2
公开(公告)日：2018-05-15
申请号：US15052702
申请日：2016-02-24
CPC分类号： G06T11/60 , G06T3/40 , G06T7/0081 , G06T7/11 , G06T2207/20104 , G06T2207/20132
摘要： Embodiments of the present invention are directed to optimizing image cropping. In accordance with some embodiments of the present invention, an image and an indication of an area of interest within the image are obtained. Thereafter, an amount to scale the image is determined based on a size of a container into which the image is to be placed for display. The amount to scale the image is greater for containers of a smaller size to focus on the area of interest within the image than the amount to scale the image for containers of a larger size. The image can be scaled in accordance with the determined amount to scale the image, and thereafter cropped to fit within the boundaries of the container.
-
公开(公告)号：US09972109B2
公开(公告)日：2018-05-15
申请号：US14814273
申请日：2015-07-30
发明人： Anant Gilra
IPC分类号： G06F3/0484 , G06T11/60
CPC分类号： G06T11/60 , G06F3/04845 , G06F3/04847 , G06T2200/24
摘要： A computing device is provided. The computing device includes a memory, at least one processor coupled to the memory and an action undo module executable by the processor. The action undo module is configured to receive user input requesting a target session of one or more sessions be undone. Each session of the one or more sessions includes at least one identifier of at least one tile that is affected by a respective action. The target session includes an identifier of a target action having affected one or more tiles identified in a collateral session of the one or more sessions. The collateral session includes an identifier of a collateral action. In one embodiment, the target session is lower in the first stack than the collateral session and the action undo module is configured to undo the collateral action and undo the target action.
-
公开(公告)号：US09965717B2
公开(公告)日：2018-05-08
申请号：US14940916
申请日：2015-11-13
发明人： Zhaowen Wang , Xianming Liu , Hailin Jin , Chen Fang
CPC分类号： G06N3/0454 , G06K9/4628 , G06K9/6257 , G06K9/627 , G06K9/628 , G06K9/6284 , G06K9/629 , G06N3/04 , G06N3/08 , G06Q50/01
摘要： Embodiments of the present invention relate to learning image representation by distilling from multi-task networks. In implementation, more than one single-task network is trained with heterogeneous labels. In some embodiments, each of the single-task networks is transformed into a Siamese structure with three branches of sub-networks so that a common triplet ranking loss can be applied to each branch. A distilling network is trained that approximates the single-task networks on a common ranking task. In some embodiments, the distilling network is a Siamese network whose ranking function is optimized to approximate an ensemble ranking of each of the single-task networks. The distilling network can be utilized to predict tags to associate with a test image or identify similar images to the test image.
-
公开(公告)号：US20180122050A1
公开(公告)日：2018-05-03
申请号：US15338952
申请日：2016-10-31
发明人： Abhishek Singh , Vivek Agrawal
IPC分类号： G06T5/00 , G06T1/20 , G06F3/0482
摘要： A digital medium environment is described to improve creation and rasterization of a shape through pixel alignment. In one example, a pixel alignment system is implemented at least partially in hardware of a computing device. The pixel alignment system receives an input that specifies a geometry, a stroke setting, and a location that serves as a basis to position the shape. The pixel alignment system then snaps the location as specified by the at least one input to a snapped location based on a pixel grid. The snapped location based on the geometry, the stroke setting, and the location as specified by the input. A rasterization module is then employed to rasterize the shape as pixels based on the snapped location.
-
公开(公告)号：US20180114097A1
公开(公告)日：2018-04-26
申请号：US15853120
申请日：2017-12-22
发明人： Zhaowen Wang , Luoqi Liu , Hailin Jin
CPC分类号： G06K9/6257 , G06K9/4628 , G06K9/6272 , G06K9/6277 , G06K9/66 , G06K9/6828 , G06K2209/01 , G06N3/0454
摘要： Font recognition and similarity determination techniques and systems are described. In a first example, localization techniques are described to train a model using machine learning (e.g., a convolutional neural network) using training images. The model is then used to localize text in a subsequently received image, and may do so automatically and without user intervention, e.g., without specifying any of the edges of a bounding box. In a second example, a deep neural network is directly learned as an embedding function of a model that is usable to determine font similarity. In a third example, techniques are described that leverage attributes described in metadata associated with fonts as part of font recognition and similarity determinations.
-
公开(公告)号：US09953425B2
公开(公告)日：2018-04-24
申请号：US14447296
申请日：2014-07-30
发明人： Zhe Lin , Hailin Jin , Jianchao Yang
CPC分类号： G06T7/33 , G06K9/627 , G06N3/0454
摘要： A first set of attributes (e.g., style) is generated through pre-trained single column neural networks and leveraged to regularize the training process of a regularized double-column convolutional neural network (RDCNN). Parameters of the first column (e.g., style) of the RDCNN are fixed during RDCNN training. Parameters of the second column (e.g., aesthetics) are fine-tuned while training the RDCNN and the learning process is supervised by the label identified by the second column (e.g., aesthetics). Thus, features of the images may be leveraged to boost classification accuracy of other features by learning a RDCNN.
-
公开(公告)号：US09942117B1
公开(公告)日：2018-04-10
申请号：US15414160
申请日：2017-01-24
发明人： Shiv Kumar Saini , Ritwik Sinha
CPC分类号： H04L43/067 , H04L41/5038
摘要： Metric anomaly detection techniques in a digital medium environment are described. An input time interval is identified by an analytics system for the presence of an anomaly of a metric. Inclusion of a scheduled event in the input time interval is identified, and a historical time interval is determined that also includes the scheduled event. Usage data describing values of the metric is then obtained for both the input time interval and the historical time interval. The usage data corresponding to the input time interval is then compared with the usage data corresponding to the historical time interval to detect effects of the scheduled event and whether the input time interval includes an anomaly in the metric.
-
公开(公告)号：US09940577B2
公开(公告)日：2018-04-10
申请号：US14793157
申请日：2015-07-07
发明人： Hailin Jin , Jonathan Krause , Jianchao Yang
CPC分类号： G06N3/088 , G06F17/30247 , G06K9/00362 , G06K9/4628 , G06K9/6218 , G06K9/627 , G06N99/005
摘要： Embodiments of the present invention relate to finding semantic parts in images. In implementation, a convolutional neural network (CNN) is applied to a set of images to extract features for each image. Each feature is defined by a feature vector that enables a subset of the set of images to be clustered in accordance with a similarity between feature vectors. Normalized cuts may be utilized to help preserve pose within each cluster. The images in the cluster are aligned and part proposals are generated by sampling various regions in various sizes across the aligned images. To determine which part proposal corresponds to a semantic part, a classifier is trained for each part proposal and semantic part to determine which part proposal best fits the correlation pattern given by the true semantic part. In this way, semantic parts in images can be identified without any previous part annotations.
-
公开(公告)号：US09939925B2
公开(公告)日：2018-04-10
申请号：US14090011
申请日：2013-11-26
发明人： Daichi Ito , Nathan Carr , Naoki Sasaki
IPC分类号： G06F3/0346 , G06F3/00 , G06F3/03 , G06F3/0481 , G06F3/0354
CPC分类号： G06F3/0346 , G06F3/005 , G06F3/0304 , G06F3/03545 , G06F3/04815
摘要： Example systems and methods of providing a user interface are presented. In one example, a graphical object is displayed on an opaque display component on a user-facing side of a computing device. Using a sensing component of the computing device, movement of a physical pointer controlled by a user is sensed. The physical pointer may be located opposite the user-facing side of the computing device. On the opaque display component, a representation of the physical pointer is displayed during the movement of the physical pointer. The graphical object, as displayed on the opaque display component, is modified based on the sensed movement of the physical pointer during the movement of the physical pointer.
-
-
-
-
-
-
-
-
-
搜索结果
5157 条记录 (耗时 0.044 秒)
