Abstract: A classification system classifies different aspects of content of an input image stream, such as faces, landmarks, events, and so forth. The classification system includes a general classifier and at least one specialized classifier template. The general classifier is trained to classify a large number of different aspects of content, and a specialized classifier can be trained based on a specialized classifier template during operation of the classification system to classify a particular subset of the multiple different aspects of content. The classification system determines when to use the general classifier and when to use a specialized classifier based on class skew, which refers to the temporal locality of a subset of aspects of content in the image stream.

Abstract: A classification system classifies different aspects of content of an input image stream, such as faces, landmarks, events, and so forth. The classification system includes a general classifier and at least one specialized classifier template. The general classifier is trained to classify a large number of different aspects of content, and a specialized classifier can be trained based on a specialized classifier template during operation of the classification system to classify a particular subset of the multiple different aspects of content. The classification system determines when to use the general classifier and when to use a specialized classifier based on class skew, which refers to the temporal locality of a subset of aspects of content in the image stream.

Abstract: A classification system classifies different aspects of content of an input image stream, such as faces, landmarks, events, and so forth. The classification system includes a general classifier and at least one specialized classifier template. The general classifier is trained to classify a large number of different aspects of content, and a specialized classifier can be trained based on a specialized classifier template during operation of the classification system to classify a particular subset of the multiple different aspects of content. The classification system determines when to use the general classifier and when to use a specialized classifier based on class skew, which refers to the temporal locality of a subset of aspects of content in the image stream.

Abstract: Embodiments include processes, systems, and devices that allow a white space base station to request available frequency ranges for white space transmission in a local area. A white space finder service models a primary user device's transmission signal propagation area using terrain data associated with the local area of the primary user device. The white space finder service also determines, based on the location of the white space base station and the modeled propagation area, one or more locally available, non-interfering frequency ranges and provides them to the white space base station. The white space base station compares the provided frequency ranges to policies and selects one or more of the available frequencies that accommodate the policies. The white space base station also maps the transmission frequency ranges to virtual frequency ranges for transmission by a software-defined radio employing spectrum virtualization.

Abstract: A classification system classifies different aspects of content of an input image stream, such as faces, landmarks, events, and so forth. The classification system includes a general classifier and at least one specialized classifier template. The general classifier is trained to classify a large number of different aspects of content, and a specialized classifier can be trained based on a specialized classifier template during operation of the classification system to classify a particular subset of the multiple different aspects of content. The classification system determines when to use the general classifier and when to use a specialized classifier based on class skew, which refers to the temporal locality of a subset of aspects of content in the image stream.

Abstract: Embodiments include processes, systems, and devices for reshaping virtual baseband signals for transmission on non-contiguous and variable portions of a physical baseband, such as a white space frequency band. In the transmission path, a spectrum virtualization layer maps a plurality of frequency components derived from a transmission symbol produced by a physical layer protocol to sub-carriers of the allocated physical frequency band. The spectrum virtualization layer then outputs a time-domain signal derived from the mapped frequency components. In the receive path, a time-domain signal received on the physical baseband is reshaped by the virtual spectrum layer in order to recompose a time-domain symbol in the virtual baseband.

Abstract: Embodiments include processes, systems, and devices for reshaping virtual baseband signals for transmission on non-contiguous and variable portions of a physical baseband, such as a white space frequency band. In the transmission path, a spectrum virtualization layer maps a plurality of frequency components derived from a transmission symbol produced by a physical layer protocol to sub-carriers of the allocated physical frequency band. The spectrum virtualization layer then outputs a time-domain signal derived from the mapped frequency components. In the receive path, a time-domain signal received on the physical baseband is reshaped by the virtual spectrum layer in order to recompose a time-domain symbol in the virtual baseband.

Abstract: Embodiments include processes, systems, and devices that allow a white space base station to request available frequency ranges for white space transmission in a local area. A white space finder service models a primary user device's transmission signal propagation area using terrain data associated with the local area of the primary user device. The white space finder service also determines, based on the location of the white space base station and the modeled propagation area, one or more locally available, non-interfering frequency ranges and provides them to the white space base station. The white space base station compares the provided frequency ranges to policies and selects one or more of the available frequencies that accommodate the policies. The white space base station also maps the transmission frequency ranges to virtual frequency ranges for transmission by a software-defined radio employing spectrum virtualization.

Abstract: Embodiments include processes, systems, and devices for reshaping virtual baseband signals for transmission on non-contiguous and variable portions of a physical baseband, such as a white space frequency band. In the transmission path, a spectrum virtualization layer maps a plurality of transmission components associated with a transmission symbol produced by a physical layer protocol to sub-carriers of the allocated physical frequency band. The spectrum virtualization layer then outputs a physical transmission symbol composed of time-domain samples derived from the mapped frequency components and a cyclic prefix. In the receive path, a time-domain symbol received on the physical baseband is reshaped and equalized by the virtual spectrum layer in order to recompose a time-domain samples of a transmission stream in the virtual baseband.

Abstract: Embodiments include processes, systems, and devices for reshaping virtual baseband signals for transmission on non-contiguous and variable portions of a physical baseband, such as a white space frequency band. In the transmission path, a spectrum virtualization layer maps a plurality of transmission components associated with a transmission symbol produced by a physical layer protocol to sub-carriers of the allocated physical frequency band. The spectrum virtualization layer then outputs a physical transmission symbol composed of time-domain samples derived from the mapped frequency components and a cyclic prefix. In the receive path, a time-domain symbol received on the physical baseband is reshaped and equalized by the virtual spectrum layer in order to recompose a time-domain samples of a transmission stream in the virtual baseband.

Abstract: Embodiments include processes, systems, and devices that allow a white space base station to request available frequency ranges for white space transmission in a local area. A white space finder service models a primary user device's transmission signal propagation area using terrain data associated with the local area of the primary user device. The white space finder service also determines, based on the location of the white space base station and the modeled propagation area, one or more locally available, non-interfering frequency ranges and provides them to the white space base station. The white space base station compares the provided frequency ranges to policies and selects one or more of the available frequencies that accommodate the policies. The white space base station also maps the transmission frequency ranges to virtual frequency ranges for transmission by a software-defined radio employing spectrum virtualization.

Abstract: Embodiments include processes, systems, and devices that allow a white space base station to request available frequency ranges for white space transmission in a local area. A white space finder service models a primary user device's transmission signal propagation area using terrain data associated with the local area of the primary user device. The white space finder service also determines, based on the location of the white space base station and the modeled propagation area, one or more locally available, non-interfering frequency ranges and provides them to the white space base station. The white space base station compares the provided frequency ranges to policies and selects one or more of the available frequencies that accommodate the policies. The white space base station also maps the transmission frequency ranges to virtual frequency ranges for transmission by a software-defined radio employing spectrum virtualization.

Abstract: Embodiments include processes, systems, and devices for reshaping virtual baseband signals for transmission on non-contiguous and variable portions of a physical baseband, such as a white space frequency band. In the transmission path, a spectrum virtualization layer maps a plurality of transmission components associated with a transmission symbol produced by a physical layer protocol to sub-carriers of the allocated physical frequency band. The spectrum virtualization layer then outputs a physical transmission symbol composed of time-domain samples derived from the mapped frequency components and a cyclic prefix. In the receive path, a time-domain symbol received on the physical baseband is reshaped and equalized by the virtual spectrum layer in order to recompose a time-domain samples of a transmission stream in the virtual baseband.

Abstract: Embodiments include processes, systems, and devices for reshaping virtual baseband signals for transmission on non-contiguous and variable portions of a physical baseband, such as a white space frequency band. In the transmission path, a spectrum virtualization layer maps a plurality of frequency components derived from a transmission symbol produced by a physical layer protocol to sub-carriers of the allocated physical frequency band. The spectrum virtualization layer then outputs a time-domain signal derived from the mapped frequency components. In the receive path, a time-domain signal received on the physical baseband is reshaped by the virtual spectrum layer in order to recompose a time-domain symbol in the virtual baseband.