Abstract: Systems and methods for detecting crop marks depicted in digital images are disclosed. According to certain aspects, an electronic device detects a set of crop marks in a reference digital image and aligns the reference digital image with an input digital image. Based on locations of the detected crop marks as aligned to the input digital image, the electronic device detects a set of partial crop marks in the input digital image. As a result, the partial crop marks in the input digital image may be removed.

Abstract: Photonic interposers that enable low-power, high-bandwidth inter-chip (e.g., board-level and/or rack-level) as well as intra-chip communication are described. Described herein are techniques, architectures and processes that improve upon the performance of conventional computers. Some embodiments provide photonic interposers that use photonic tiles, where each tile includes programmable photonic circuits that can be programmed based on the needs of a particular computer architecture. Some tiles are instantiations of a common template tile that are stitched together in a 1D or a 2D arrangement. Some embodiments described herein provide a programmable physical network designed to connect pairs of tiles together with photonic links.

Abstract: Systems and methods for automatically determining and displaying salient portions of images are disclosed. According to certain aspects, an electronic device may support a design application that may apply a saliency detection learning model to a digital image, resulting in the application generating one or more salient portions of the digital image. The electronic device may generate a digital rendering of the salient portion of the image on digital models of items or products, and may enable a user to review the digital rendering. The user may also choose alternative salient portions of the digital image and/or aspect ratios for those salient portions for inclusion on a digital model of the item or product.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for retrieving and using contextual data from previous conversation sessions in conversational searches. In one aspect, a method includes receiving a first query for a first user session, determining that the first query refers to one or more tags in a first repository, the first repository associating respective identifiers to respective tags, each identifier representing a corresponding user session, determining one or more particular identifiers associated with the one or more tags in the first repository, retrieving particular contextual data associated with the determined particular identifiers in a second repository, the second repository associating respective identifiers to respective contextual data associated with corresponding user sessions represented by the respective identifiers, and performing an action responsive to the first query based on the retrieved particular contextual data.

Abstract: Systems and methods for automatically determining and displaying salient portions of images are disclosed. According to certain aspects, an electronic device may support a design application that may apply a saliency detection learning model to a digital image, resulting in the application generating one or more salient portions of the digital image. The electronic device may generate a digital rendering of the salient portion of the image on digital models of items or products, and may enable a user to review the digital rendering. The user may also choose alternative salient portions of the digital image and/or aspect ratios for those salient portions for inclusion on a digital model of the item or product.

Abstract: Aspects of the disclosure provide for the generation of a driving difficulty heat map for autonomous vehicles. For instance, log data generated by a vehicle being driven in a manual driving mode for a segment of a route may be input into a disengage model in order to generate an output identifying a likelihood of a vehicle driving in an autonomous driving mode requiring a disengage from the autonomous driving mode along the segment of the route. The log data may have been collected within a geographic area. A grid for the geographic area may be generated. The grid may include a plurality of cells. The output is assigned to one of the plurality of cells. The plurality of cells and assigned output may be used to generate a driving difficulty heat map for the geographic area.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for modifying queries in a discourse context. In one aspect, a method includes determining whether a query received from a user device subsequent to information provided to a user device and that defines a discourse context belongs to the discourse context. If the query belongs to the discourse context, the query is modified to take into account attributes of the discourse context.

Abstract: Aspects of the disclosure provide for the selection of a route for a vehicle having an autonomous driving mode. For instance, an initial location of the vehicle may be identified. This location may be used to determine a set of possible routes to a destination location. A cost for each route of the plurality is determined by inputting time of day information, map information, and details of that route into one or more models in order to determine whether the vehicle is likely to be stranded along that route and assessing the cost based at least in part on the determination of whether the vehicle is likely to be stranded along that route. One of the routes of the set of possible routes may be selected based on any determined costs. The vehicle may be controlled in the autonomous driving mode using the selected one.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for retrieving and using contextual data from previous conversation sessions in conversational searches. In one aspect, a method includes receiving a first query for a first user session, determining that the first query refers to one or more tags in a first repository, the first repository associating respective identifiers to respective tags, each identifier representing a corresponding user session, determining one or more particular identifiers associated with the one or more tags in the first repository, retrieving particular contextual data associated with the determined particular identifiers in a second repository, the second repository associating respective identifiers to respective contextual data associated with corresponding user sessions represented by the respective identifiers, and performing an action responsive to the first query based on the retrieved particular contextual data.

Abstract: Provided herein are optical flow switches, and optical flow switch packages, implemented using photonic switches. The optical flow switch includes a network of photonic switches arranged between input and output ports of the optical flow switch. The network of photonic switches spans two or more reticles, and the two or more reticles may include photonic switching arrangements corresponding to repeated reticle masks or sets of reticle masks. The optical flow switch may be mounted to a glass substrate to form an optical flow switch package.

Abstract: Aspects of the disclosure provide for the determining whether to assign a driver to a trip. For instance, a route for a trip between a pickup location and a destination location for a passenger may be determined. Branch routes for the route may be determined based on one or more lane changes of the route. A probability of following each determined branch route may be determined. A probability of requiring assistance for each determined branch route may be determined. Whether to assign a driver to the trip may be determined based on the probabilities for each determined branch route. A vehicle of a fleet of vehicles may be assigned to the trip based on the determine of whether to assign a driver.

Abstract: Photonic interposers that enable low-power, high-bandwidth inter-chip (e.g., board-level and/or rack-level) as well as intra-chip communication are described. Described herein are techniques, architectures and processes that improve upon the performance of conventional computers. Some embodiments provide photonic interposers that use photonic tiles, where each tile includes programmable photonic circuits that can be programmed based on the needs of a particular computer architecture. Some tiles are instantiations of a common template tile that are stitched together in a 1D or a 2D arrangement. Some embodiments described herein provide a programmable physical network designed to connect pairs of tiles together with photonic links.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for retrieving and using contextual data from previous conversation sessions in conversational searches. In one aspect, a method includes receiving a first query for a first user session, determining that the first query refers to one or more tags in a first repository, the first repository associating respective identifiers to respective tags, each identifier representing a corresponding user session, determining one or more particular identifiers associated with the one or more tags in the first repository, retrieving particular contextual data associated with the determined particular identifiers in a second repository, the second repository associating respective identifiers to respective contextual data associated with corresponding user sessions represented by the respective identifiers, and performing an action responsive to the first query based on the retrieved particular contextual data.

Abstract: The disclosure relates to assessing impact of blockages on an autonomous vehicle transportation service. An example method may include identifying a plurality of starting and destination location pairs within a service area; running a first plurality of simulations by determining a first route for each of the plurality of pairs without including a first blockage in the service area; determining a first summary for each determined first route identifying a length of time to complete that determined route; running a second plurality of simulations by determining a second route for each of the plurality of pairs and including the first blockage in the service area; determining a second summary for each determined second route identifying a length of time to complete that determined route; and comparing the determined first summaries with the determined second summaries in order to assess an impact of the first blockage on the service.

Abstract: Photonic interposers that enable low-power, high-bandwidth inter-chip (e.g., board-level and/or rack-level) as well as intra-chip communication are described. Described herein are techniques, architectures and processes that improve upon the performance of conventional computers. Some embodiments provide photonic interposers that use photonic tiles, where each tile includes programmable photonic circuits that can be programmed based on the needs of a particular computer architecture. Some tiles are instantiations of a common template tile that are stitched together in a 1D or a 2D arrangement. Some embodiments described herein provide a programmable physical network designed to connect pairs of tiles together with photonic links.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for retrieving and using contextual data from previous conversation sessions in conversational searches. In one aspect, a method includes receiving a first query for a first user session, determining that the first query refers to one or more tags in a first repository, the first repository associating respective identifiers to respective tags, each identifier representing a corresponding user session, determining one or more particular identifiers associated with the one or more tags in the first repository, retrieving particular contextual data associated with the determined particular identifiers in a second repository, the second repository associating respective identifiers to respective contextual data associated with corresponding user sessions represented by the respective identifiers, and performing an action responsive to the first query based on the retrieved particular contextual data.

Abstract: Aspects of the disclosure provide for the generation of a driving difficulty heat map for autonomous vehicles. For instance, log data generated by a vehicle being driven in a manual driving mode for a segment of a route may be input into a disengage model in order to generate an output identifying a likelihood of a vehicle driving in an autonomous driving mode requiring a disengage from the autonomous driving mode along the segment of the route. The log data may have been collected within a geographic area. A grid for the geographic area may be generated. The grid may include a plurality of cells. The output is assigned to one of the plurality of cells. The plurality of cells and assigned output may be used to generate a driving difficulty heat map for the geographic area.

Abstract: Aspects of the disclosure provide a method of providing a destination to an autonomous vehicle in order to enable the autonomous vehicle to collect data according to a targeted driving goal. For instance, a current location of an autonomous vehicle may be received. A set of destinations may be selected from a plurality of predetermined destinations. A route may be determined for each destination. A relevance score may be determined for each destination based on the determined routes and the targeted driving goal. Each destination may be assigned to one of a set of two or more buckets based on the relevance scores. A destination of the set may be selected based on a predetermined sampling probability. The selected destination is sent to the autonomous vehicle in order to cause the autonomous vehicle to travel to the selected destination in an autonomous driving mode.

Abstract: Systems and methods for automatically determining and displaying salient portions of images are disclosed. According to certain aspects, an electronic device may support a design application that may apply a saliency detection learning model to a digital image, resulting in the application generating one or more salient portions of the digital image. The electronic device may generate a digital rendering of the salient portion of the image on digital models of items or products, and may enable a user to review the digital rendering. The user may also choose alternative salient portions of the digital image and/or aspect ratios for those salient portions for inclusion on a digital model of the item or product.

Abstract: Systems and methods for detecting crop marks depicted in digital images are disclosed. According to certain aspects, an electronic device detects a set of crop marks in a reference digital image and aligns the reference digital image with an input digital image. Based on locations of the detected crop marks as aligned to the input digital image, the electronic device detects a set of partial crop marks in the input digital image. As a result, the partial crop marks in the input digital image may be removed.