Abstract: Incorporation of an exogenous large-vocabulary model into rule-based speech recognition is provided. An audio stream is received by a local small-vocabulary rule-based speech recognition system (SVSRS), and is streamed to a large-vocabulary statistically-modeled speech recognition system (LVSRS). The SVSRS and LVSRS perform recognitions of the audio. If a portion of the audio is not recognized by the SVSRS, a rule is triggered that inserts a mark-up in the recognition result. The recognition result is sent to the LVSRS. If a mark-up is detected, recognition of a specified portion of the audio is performed. The LVSRS result is unified with the SVSRS result and sent as a hybrid response back to the SVSRS. If the hybrid-recognition rule is not triggered, an arbitration algorithm is evoked to determine whether the SVSRS or the LVSRS recognition has a lesser word error rate. The determined recognition is sent as a response to the SVSRS.

Abstract: Incorporation of an exogenous large-vocabulary model into rule-based speech recognition is provided. An audio stream is received by a local small-vocabulary rule-based speech recognition system (SVSRS), and is streamed to a large-vocabulary statistically-modeled speech recognition system (LVSRS). The SVSRS and LVSRS perform recognitions of the audio. If a portion of the audio is not recognized by the SVSRS, a rule is triggered that inserts a mark-up in the recognition result. The recognition result is sent to the LVSRS. If a mark-up is detected, recognition of a specified portion of the audio is performed. The LVSRS result is unified with the SVSRS result and sent as a hybrid response back to the SVSRS. If the hybrid-recognition rule is not triggered, an arbitration algorithm is evoked to determine whether the SVSRS or the LVSRS recognition has a lesser word error rate. The determined recognition is sent as a response to the SVSRS.

Abstract: Incorporation of an exogenous large-vocabulary model into rule-based speech recognition is provided. An audio stream is received by a local small-vocabulary rule-based speech recognition system (SVSRS), and is streamed to a large-vocabulary statistically-modeled speech recognition system (LVSRS). The SVSRS and LVSRS perform recognitions of the audio. If a portion of the audio is not recognized by the SVSRS, a rule is triggered that inserts a mark-up in the recognition result. The recognition result is sent to the LVSRS. If a mark-up is detected, recognition of a specified portion of the audio is performed. The LVSRS result is unified with the SVSRS result and sent as a hybrid response back to the SVSRS. If the hybrid-recognition rule is not triggered, an arbitration algorithm is evoked to determine whether the SVSRS or the LVSRS recognition has a lesser word error rate. The determined recognition is sent as a response to the SVSRS.

Abstract: A secure transcoder assembly features a secure data path where the related crypting and coding functions are secured in the secure data path. More specifically, during transcoding the ingress (received) encrypted video data is decrypted and copied from application memory (that is accessible to the application processor) to secure memory (that is inaccessible to the application processor) by security processor(s). Once in the secured memory, this video data is transcoded. The video data, now in its transcoded form, is re-encrypted by the security processor(s) during the transfer back to application memory for egress (continued transmission). The mode for the re-encrypting may be managed by the application processor via a crypto API, and the application processor may further feature an opaque handle on the data when it is passed to the secure region.

Abstract: Incorporation of an exogenous large-vocabulary model into rule-based speech recognition is provided. An audio stream is received by a local small-vocabulary rule-based speech recognition system (SVSRS), and is streamed to a large-vocabulary statistically-modeled speech recognition system (LVSRS). The SVSRS and LVSRS perform recognitions of the audio. If a portion of the audio is not recognized by the SVSRS, a rule is triggered that inserts a mark-up in the recognition result. The recognition result is sent to the LVSRS. If a mark-up is detected, recognition of a specified portion of the audio is performed. The LVSRS result is unified with the SVSRS result and sent as a hybrid response back to the SVSRS. If the hybrid-recognition rule is not triggered, an arbitration algorithm is evoked to determine whether the SVSRS or the LVSRS recognition has a lesser word error rate. The determined recognition is sent as a response to the SVSRS.

Abstract: Methods are disclosed for measuring ad impressions and receiving feedback on local ad assets inserted into a video transport stream at the set top box level. Each set top box stores the number of times an ad asset is inserted into an ad avail, along with a variety of other information relating to the playback of the ad asset. This measurement data is aggregated and sent to the ad decision service. In order to balance bandwidth usage, each set top box may report its measurement data to the ad decision service at a different time interval that is randomly selected. As it is desirable to receive the data in a timely manner, the random intervals may be confined so that all measurement data is reported within a predefined time period, such as for example over a twelve hour period.

Abstract: Methods are disclosed for inserting local ad assets into a video transport stream at the set top box level. The media advertising platform of the present system works in conjunction with existing platforms, such as an advertising decision service and a media platform. The present system further includes a client resident on end user set top boxes. In general, the present system operates by pre-caching advertisements to a set top box or boxes within a household or elsewhere. Each set top box is also assigned to particular groups, based on characteristics of the user of that set top box. Group membership information is also sent to the set top box and stored. The present system sends the set top box a decision matrix based on group memberships to allow the set top box to select and insert an ad asset targeted to the specific set top box.

Abstract: A secure transcoder assembly features a secure data path where the related crypting and coding functions are secured in the secure data path. More specifically, during transcoding the ingress (received) encrypted video data is decrypted and copied from application memory (that is accessible to the application processor) to secure memory (that is inaccessible to the application processor) by security processor(s). Once in the secured memory, this video data is transcoded. The video data, now in its transcoded form, is re-encrypted by the security processor(s) during the transfer back to application memory for egress (continued transmission). The mode for the re-encrypting may be managed by the application processor via a crypto API, and the application processor may further feature an opaque handle on the data when it is passed to the secure region.

Abstract: Methods are disclosed for measuring ad impressions and receiving feedback on local ad assets inserted into a video transport stream at the set top box level. Each set top box stores the number of times an ad asset is inserted into an ad avail, along with a variety of other information relating to the playback of the ad asset. This measurement data is aggregated and sent to the ad decision service. In order to balance bandwidth usage, each set top box may report its measurement data to the ad decision service at a different time interval that is randomly selected. As it is desirable to receive the data in a timely manner, the random intervals may be confined so that all measurement data is reported within a predefined time period, such as for example over a twelve hour period.

Abstract: Methods are disclosed for inserting local ad assets into a video transport stream at the set top box level. The media advertising platform of the present system works in conjunction with existing platforms, such as an advertising decision service and a media platform. The present system further includes a client resident on end user set top boxes. In general, the present system operates by pre-caching advertisements to a set top box or boxes within a household or elsewhere. Each set top box is also assigned to particular groups, based on characteristics of the user of that set top box. Group membership information is also sent to the set top box and stored. The present system sends the set top box a decision matrix based on group memberships to allow the set top box to select and insert an ad asset targeted to the specific set top box.

Abstract: Various embodiments pertain to methods and systems for providing multi-source content in electronic program guides (EPGs). In at least some embodiments, various techniques can present the viewer with a user interface, such as a grid, that is efficiently and intelligently assembled and managed to foster a desirable user experience.

Abstract: Systems and methods for repositioning selectable content. Content that includes selectable items is displayed on a display device through a user interface. When an item is selected, the item is repositioned and sometimes resized so that the selected item is more easily viewed. If the item includes text, the text is more legible because it has been enlarged. Other items included in the content that are not currently selected are not enlarged or modified. When a new item is selected, the new item is repositioned or resized and the previously selected item returns to a previous state.

Abstract: Systems and methods for repositioning selectable content. Content that includes selectable items is displayed on a display device through a user interface. When an item is selected, the item is repositioned and sometimes resized so that the selected item is more easily viewed. If the item includes text, the text is more legible because it has been enlarged. Other items included in the content that are not currently selected are not enlarged or modified. When a new item is selected, the new item is repositioned or resized and the previously selected item returns to a previous state.