Abstract: The present disclosure provides methods and apparatuses for evaluating a quality of an image detected by a camera of a video conferencing device. A method includes sampling a current frame of an input video stream, extracting image quality information from the current frame, comparing the extracted image quality information with reference image quality information generated by an image quality model, selecting, based on the comparing, an image quality mode of the current frame, and proceeding with performing image analysis on the current frame, based on the image quality mode. The image analysis includes at least one of face recognition and object recognition.

Abstract: Disposed are a method for preparing cellulose-polyvinyl alcohol fabric with improved surface characteristics, cellulose-polyvinyl alcohol fabric with improved surface characteristics prepared thereby, and fabric for wound healing comprising same. The method for preparing cellulose-polyvinyl alcohol fabric with improved surface characteristics comprises, after preparing a composite including a cellulose-based polymer and a polyvinyl alcohol-based polymer, exposing the composite to an alkali solution. The cellulose-polyvinyl alcohol fabric with improved surface characteristics is prepared using the method for preparing cellulose-polyvinyl alcohol fabric with improved surface characteristics. The fabric for wound healing includes the cellulose-polyvinyl alcohol fabric with improved surface characteristics.

Abstract: A novelty detector incudes a generator configured to output reconstructed data from actual data; and a discriminator configured to receive the actual data as well as the reconstructed data and to produce, using the actual data and the reconstructed data, discrimination data representing whether the actual data is normal or abnormal.

Abstract: The present disclosure provides methods and apparatuses for evaluating a quality of an image detected by a camera of a video conferencing device. A method includes sampling a current frame of an input video stream, extracting image quality information from the current frame, comparing the extracted image quality information with reference image quality information generated by an image quality model, selecting, based on the comparing, an image quality mode of the current frame, and proceeding with performing image analysis on the current frame, based on the image quality mode. The image analysis includes at least one of face recognition and object recognition.

Abstract: A novelty detector incudes a generator configured to output reconstructed data from actual data; and a discriminator configured to receive the actual data as well as the reconstructed data and to produce, using the actual data and the reconstructed data, discrimination data representing whether the actual data is normal or abnormal.