Abstract: A computing system generates a query that references an entity based upon an ontology of a knowledge graph and a query pattern and identifies at least one passage from amongst a plurality of passages stored in a passage repository based upon the query and at least one ranking model. The computing system identifies potential answers to the query based upon the at least one passage, the query, and a machine reading comprehension model. The computing system suppresses invalid answers in the potential answers to the query using a plurality of computer-implemented techniques, thereby identifying an answer to the query. The computing system generates a fact for the entity based upon the answer and the ontology and adds the fact to the knowledge graph such that the fact is linked to the entity in the knowledge graph, where the fact is available for querying.

Abstract: Various new and non-obvious apparatus and methods for using frame caching to improve packet loss recovery are disclosed. One of the disclosed embodiments is a method for using periodical and synchronized frame caching within an encoder and its corresponding decoder. When the decoder discovers packet loss, it informs the encoder which then generates a frame based on one of the shared frames stored at both the encoder and the decoder. When the decoder receives this generated frame it can decode it using its locally cached frame.

Abstract: Various new and non-obvious apparatus and methods for using frame caching to improve packet loss recovery are disclosed. One of the disclosed embodiments is a method for using periodical and synchronized frame caching within an encoder and its corresponding decoder. When the decoder discovers packet loss, it informs the encoder which then generates a frame based on one of the shared frames stored at both the encoder and the decoder. When the decoder receives this generated frame it can decode it using its locally cached frame.

Abstract: In one implementation, a computer-implemented method for determining user interests includes receiving user activity data that describes an interaction between a user and digital content. The method further includes generating first interest-level data that quantifies a first level of interest between the user and an entity that is a topic of the digital content, where the entity is of a plurality of entities included within a knowledge base. The knowledge base includes an ontology that indicates relationships between the plurality of entities. The method further includes identifying a candidate entity based on the candidate entity having one of the relationships to the entity within the knowledge base. Additionally, the method includes generating second interest-level data that quantifies a second level of interest between the user and the candidate entity based on an analysis of the one of the relationships of the candidate entity to the entity within the knowledge base.

Abstract: Aspects of the technology are directed to systems and methods for mitigating inconsistencies in a knowledge base. An inconsistency is automatically detected and it is determined whether the inconsistency is based on a source error, such as bad data quality, or an over conflation error of an entity. If the inconsistency is based on a source error, the inconsistent data point is removed. If the inconsistency is based on an over conflation of an entity, the entity is split up into two separate entities.

Abstract: Various new and non-obvious apparatus and methods for using frame caching to improve packet loss recovery are disclosed. One of the disclosed embodiments is a method for using periodical and synchronized frame caching within an encoder and its corresponding decoder. When the decoder discovers packet loss, it informs the encoder which then generates a frame based on one of the shared frames stored at both the encoder and the decoder. When the decoder receives this generated frame it can decode it using its locally cached frame.

Abstract: Various new and non-obvious apparatus and methods for using frame caching to improve packet loss recovery are disclosed. One of the disclosed embodiments is a method for using periodical and synchronized frame caching within an encoder and its corresponding decoder. When the decoder discovers packet loss, it informs the encoder which then generates a frame based on one of the shared frames stored at both the encoder and the decoder. When the decoder receives this generated frame it can decode it using its locally cached frame.

Abstract: In one implementation, a computer-implemented method for determining user interests includes receiving user activity data that describes an interaction between a user and digital content. The method further includes generating first interest-level data that quantifies a first level of interest between the user and an entity that is a topic of the digital content, where the entity is of a plurality of entities included within a knowledge base. The knowledge base includes an ontology that indicates relationships between the plurality of entities. The method further includes identifying a candidate entity based on the candidate entity having one of the relationships to the entity within the knowledge base. Additionally, the method includes generating second interest-level data that quantifies a second level of interest between the user and the candidate entity based on an analysis of the one of the relationships of the candidate entity to the entity within the knowledge base.

Abstract: Techniques and tools are described for flexible range reduction of samples of video. For example, an encoder signals a first set of one or more syntax elements for range reduction of luma samples and signals a second set of one or more syntax elements for range reduction of chroma samples. The encoder selectively scales down the luma samples and chroma samples in a manner consistent with the first syntax element(s) and second syntax element(s), respectively. Or, an encoder signals range reduction syntax element(s) in an entry point header for an entry point segment, where the syntax element(s) apply to pictures in the entry point segment. If range reduction is used for the pictures, the encoder scales down samples of the pictures. Otherwise, the encoder skips the scaling down. A decoder performs corresponding parsing and scaling up operations.

Abstract: Various new and non-obvious apparatus and methods for using frame caching to improve packet loss recovery are disclosed. One of the disclosed embodiments is a method for using periodical and synchronized frame caching within an encoder and its corresponding decoder. When the decoder discovers packet loss, it informs the encoder which then generates a frame based on one of the shared frames stored at both the encoder and the decoder. When the decoder receives this generated frame it can decode it using its locally cached frame.

Abstract: Various new and non-obvious apparatus and methods for using frame caching to improve packet loss recovery are disclosed. One of the disclosed embodiments is a method for using periodical and synchronized frame caching within an encoder and its corresponding decoder. When the decoder discovers packet loss, it informs the encoder which then generates a frame based on one of the shared frames stored at both the encoder and the decoder. When the decoder receives this generated frame it can decode it using its locally cached frame.

Abstract: Techniques and tools are described for adaptive deadzone (“DZ”) resizing during quantization. For example, in some embodiments, for quantization of an AC frequency coefficient of a block, a video encoder adjusts DZ size of a selected quantizer depending on the texture of the block. In other embodiments, a video encoder adjusts DZ size depending on the frequency of a coefficient being quantized. In still other embodiments, for quantization of an AC frequency coefficient of a block, a video encoder adjusts DZ size depending on the texture of the block and the frequency of the coefficient being quantized.

Abstract: A video encoder includes a region detector module that classifies blocks of video frames. An adaptive filter module applies a median filter to a block based upon a block classification assigned by the region detector module. An adaptive quantization module quantizes a block according to a quantization method adaptively determined based upon a block classification assigned by the region detection module. In one example, a video encoder adaptively determines a median filter selected using a block classification. In another example, a video encoder adaptively determines whether to drop an isolated last transform coefficient based on the block classification, and/or applies a dead-zone selected using the block classification.

Abstract: Various new and non-obvious apparatus and methods for using frame caching to improve packet loss recovery are disclosed. One of the disclosed embodiments is a method for using periodical and synchronized frame caching within an encoder and its corresponding decoder. When the decoder discovers packet loss, it informs the encoder which then generates a frame based on one of the shared frames stored at both the encoder and the decoder. When the decoder receives this generated frame it can decode it using its locally cached frame.

Abstract: Techniques and tools are described for flexible range reduction of samples of video. For example, an encoder signals a first set of one or more syntax elements for range reduction of luma samples and signals a second set of one or more syntax elements for range reduction of chroma samples. The encoder selectively scales down the luma samples and chroma samples in a manner consistent with the first syntax element(s) and second syntax element(s), respectively. Or, an encoder signals range reduction syntax element(s) in an entry point header for an entry point segment, where the syntax element(s) apply to pictures in the entry point segment. If range reduction is used for the pictures, the encoder scales down samples of the pictures. Otherwise, the encoder skips the scaling down. A decoder performs corresponding parsing and scaling up operations.

Abstract: Techniques and tools are described for flexible range reduction of samples of video. For example, an encoder signals a first set of one or more syntax elements for range reduction of luma samples and signals a second set of one or more syntax elements for range reduction of chroma samples. The encoder selectively scales down the luma samples and chroma samples in a manner consistent with the first syntax element(s) and second syntax element(s), respectively. Or, an encoder signals range reduction syntax element(s) in an entry point header for an entry point segment, where the syntax element(s) apply to pictures in the entry point segment. If range reduction is used for the pictures, the encoder scales down samples of the pictures. Otherwise, the encoder skips the scaling down. A decoder performs corresponding parsing and scaling up operations.

Abstract: The present application is directed to techniques and tools for adapting the quantizer as well as the quantization step size used during video encoding. For example, a video encoder uses a non-uniform quantizer (having a relatively large dead zone ratio) for B-picture types, but uses a uniform quantizer for other types of pictures. In many encoding scenarios, this helps the encoder control bit rate for the B-picture types instead of resorting to higher quantization step sizes, which improves overall quality for the B-picture types.

Abstract: For each of plural blocks (e.g., 8×16 blocks) in a source interlaced video frame, a video encoder measures vertical intensity level fluctuation between neighboring lines and between alternating lines in the block, and determines an interlace effect status value for the block. For example, the encoder calculates a ratio between the fluctuations of neighboring and alternating lines and compares the ratio with one or more thresholds. The encoder also can measure horizontal intensity level fluctuation and compare it with vertical intensity fluctuation. The encoder determines a frame-level interlace coding mode (e.g., frame mode or field mode) for the frame based at least in part on the interlace effect status values. The encoder can represent a frame-level interlace coding mode decision with a bitstream element, and a decoder can perform one or more video decoding functions based at least in part on the value of the bitstream element.

Abstract: Techniques and tools are described for adaptive deadzone (“DZ”) resizing during quantization. For example, in some embodiments, for quantization of an AC frequency coefficient of a block, a video encoder adjusts DZ size of a selected quantizer depending on the texture of the block. In other embodiments, a video encoder adjusts DZ size depending on the frequency of a coefficient being quantized. In still other embodiments, for quantization of an AC frequency coefficient of a block, a video encoder adjusts DZ size depending on the texture of the block and the frequency of the coefficient being quantized.

Abstract: A model stored in a memory accessible by a video transcoder includes a first rate-distortion function modeling a requantization of an input video. A second-rate distortion function models a resynchronization marker insertion rate for the transcoded video, and a third rate-distortion function models an intra-block insertion rate for the transcoded video.