Cyberlink Corporation

Perfect Corp.

Ho-Chao Huang

Ming-Jun Chen

Chia-Chen Kuo

Chih-Chao Ma

Yueh-Hsuan Chiang

Po-Yu Huang

Young-Wei Lei

Hao-Ping Hung

Chung-Yi Weng

Wei-Hsin Tseng

Chen-Wei Chou

Shih-Min Tang

Huan-Wen Hsiao

Jau Hsiung Huang

Chen-Hsiu Huang

Chia-En Chiang

Mei-Chun Ku

Cheng-da Chung

Bor-Ching Lin

Jiun-Fu Chen

A computing device with a digital camera obtains a reference image depicting at least one reference color and calibrates parameters of the digital camera based on the at least one reference color. The computing device captures, by the digital camera, a digital image of an individual utilizing the calibrated parameters. The computing device defines a region of interest in a facial region of the individual depicted in the digital image captured by the digital camera. The computing device generates a skin tone profile for pixels within the region of interest and displays a predetermined makeup product recommendation based on the skin tone profile.

Systems and methods for generating skin tone profiles

In one embodiment, an image editing method, comprising: defining a region as a foreground object and a remaining region as background; applying an image warping process to a predefined object region, wherein a distortion to at least the remaining region occurs based on the image warping process applied to the predefined object region; responsive to the image warping process, determining a region to be repaired, the region to be repaired comprising a missing background; repairing the region to be repaired based on an inpainting process; and generating a repaired image by combining the predefined object region to which the image warping process is applied, one or more non-distorted regions, and the repaired region.

Image editing method and system

A method for tracking an object in a plurality of frames includes obtaining a contour of an object in a frame. For each remaining frame among the plurality of frames, the following steps are performed. A probability map generator generates a plurality of probability maps containing probability values for one or more pixels for a current frame. A contour is estimated for the current frame based on the plurality of probability maps. User input is obtained for refining the estimated contour. Based on the user input, one or more local regions to be added and/or removed are identified and recorded in a local region list. The local region list is maintained and updated during the tracking process and is used together with the probability map generator.

Systems and methods for object tracking based on user refinement input

Various embodiments are disclosed for image editing. A frame is obtained from a frame sequence depicting at least one individual, and facial characteristics in the frame are analyzed. A utilization score is assigned to the frame based on the detected facial characteristics, and a determination of whether to utilize the frame is made based on the utilization score. A completeness value is assigned, and a determination is made based on the completeness value of whether to repeat the steps above for an additional frame in the frame sequence based on the completeness value. Regions from the frames are combined to generate a composite image.

Systems and methods for image editing

Various embodiments are disclosed for selecting a region of interest within an object. One embodiment is a method for editing a digital image in an image editing device. The method comprises obtaining a first position within a digital image, the first position being specified by a pointer. The method further comprises defining a first selection region based on the first position and obtaining a second position within the digital image, the second position being specified by the pointer. The method further comprises determining movement of the pointer between the first position and the second position and predicting future movement by the pointer based on the determined movement. Based on the determined movement and the predicted future movement, the first selection region is expanded to define a second selection region.

Systems and methods for performing object selection

Various embodiments are disclosed for performing inpainting. One embodiment is a method for editing a digital image in an image editing device. The method comprises obtaining an inpainting region in the digital image, determining a target resolution for scaling a resolution of the digital image based on an original resolution of the digital image, and determining an intermediate resolution level for scaling a resolution of the digital image based on the target resolution. The method further comprises scaling the resolution of the digital image to the intermediate resolution level, performing partial inpainting of the inpainting region at the intermediate resolution, and performing inpainting on a remainder portion in the inpainting region at a final target resolution.

Systems and methods for multi-resolution inpainting

Various embodiments are disclosed for performing image inpainting. One embodiment is a method for editing a digital image in an image editing device. The method comprises obtaining a restoration region in the digital image and determining texture information corresponding to the restoration region. Based on the texture information, a texture strength map is generated corresponding to the restoration region. Based on the texture strength map, an inpainting operation is applied to pixels in the restoration region.

Systems and methods for performing image inpainting based on texture analysis

Various embodiments are disclosed for performing image inpainting. One embodiment is a method for editing a digital image in an image editing device that comprises obtaining a restoration region in the digital image and generating a structure strength map corresponding to the restoration region based on structure characteristics associated with each pixel in the restoration region. Based on the structure strength map, priority levels are determined for pixels in the restoration region. An inpainting operation is applied to the pixels in the restoration region, beginning with a pixel having a highest relative priority determined based on the structure characteristics.

Systems and methods for performing image inpainting

Disclosed are systems, methods, and computer readable media having programs for classifying sports video. In one embodiment, a method includes: extracting, from an audio stream of a video clip, a plurality of key audio components contained therein; and classifying, using at least one of the plurality of key audio components, a sport type contained in the video clip. In one embodiment, a computer readable medium having a computer program for classifying ports video includes: logic configured to extract a plurality of key audio components from a video clip; and logic configured to classify a sport type corresponding to the video clip.

Systems and methods for classifying sports video

The present invention discloses a method of processing a sports video. The method includes the steps of: analyzing the sports video to detect at least one semantic event and each of the at least one semantic event associated with a segment length; and generating a summarized video according to the detected at least one semantic event.

Method for processing a sports video and apparatus thereof

Disclosed are systems and methods for generating and embedding scene processing information in a multimedia source. Scene processing information is generated based on an analysis of audio, video and data properties of the multimedia source. The embedded scene processing information can be extracted from the multimedia source to assist in generating playback of the multimedia source as well as generating a summarized playback of the multimedia source.

Systems and method for embedding scene processing information in a multimedia source using an importance value

Systems and methods for generating a summary of photos from a plurality of received photos are described. The received photos are classified according to predefined attributes. Two or more of the categories are selected, and a ratio value is received from a user relating to the two or more of the categories. Photos are selected from among the photos in the two or more categories based on the specified ratio and based on sorting the received photos according to time information. The selected photos comprising the summary of photos are displayed.

Systems and methods for summarizing photos based on photo information and user preference

In a method for determining highlight segments of a sport video, the sport video having a plurality of image frames is analyzed to recognize a scoreboard image and a plurality of scoreboard characteristics thereof. The scoreboard image comprises at least one of the scoreboard characteristics. Then, a request message showing at least one recognized scoreboard characteristic is displayed on a user interface to receive a user response, wherein the user response includes a recognized result of one of the scoreboard characteristics. Finally, variations in the recognized scoreboard image among image frames can be monitored according to the user response. Therefore, the highlight segments of the sport video can be determined according to the variations.

Method and apparatus for determining highlight segments of sport video

A method capable of detecting a scoreboard in a program includes detecting a region which displays static images in a program having a plurality of frames, identifying the text contents of the region, and determining whether the variation in the text contents is larger than a predetermined value and occurs conforming to a predetermined rule. If the variation in the text contents is larger than the predetermined value and occurs conforming to the predetermined rule, the method signifies that the region displays the static images corresponding to a scoreboard.

Method and display system capable of detecting a scoreboard in a program

A method for re-orientating digital images in an image-editing environment where images are loaded in landscape mode by default, with the aim of unburdening users from the inconvenience of manual correction of image orientation. Intelligent re-orientation of digital images is realized by analyzing an image in order to determine if re-orientation is required, using image zone analysis processes and facial feature analysis and, by the information provided by the same processes, also determining the direction and scale of correction need to render an image correctly orientated with respect to a user.

Growing community of inventors

Taipei, Taiwan

Ho-Chao Huang