Won Young Yoo

Smoodi: Mood-based music recommendation player

In this paper, we present a mood-based music recommendation player: Smoodi. The Smoodi provides smart mood recommendation by touch and drag and it has three different views: Mood Square, Cover Flow and Mood Cloud. In the Mood Square, users can check the mood distribution of local clips and generate playlists by touching mood cells. In the Cover Flow, users can find music information and generate similar songs list by dragging a seed song. In the Tag Cloud, users can make playlists by selecting mood tags. For this application, we developed new mood model from collected mood tags and arousal-valence position values and designed a regression function to estimate mood probabilities.

A visual feature based video identifying system for the TV commercial's monitoring

For companies, TV commercials are very important way to introduce and advertise their products. It is expensive to put an advertisement on TV. So these companies generally charge other companies to verify that their TV commercials are actually broadcasted as contracted. Currently, these monitoring are done manually. All the TV programs and their air-times are recorded at the same time. Then an agent checks the start time and the end time of a TV-commercial. Video retrieval and matching techniques can be used to monitor TV commercials automatically. By extracting visual identified features, we can measure the similarities and identify a target video from thousands of videos. In this paper, we designed the visual feature DB and implemented video matching system. To construct the DB, we extracted scene change information and dominant colors of commercial clips. By combining the features, we can identify TV commercials

Real-Time monitoring system for TV commercials using video features

For companies, TV commercial is a very important way to introduce and advertise their products. It is expensive to put an advertisement on TV. So these companies generally charge other companies to monitor that their TV commercials are broadcasted properly as contracted. Currently, these monitorings have been done manually. The monitoring company records all the TV programs and their air-times while they are being broadcasted. Then the agent checks the starting-times and the ending-times of TV commercials. Video retrieval and matching techniques can be used to monitor TV commercials automatically. By extracting visual features that can identify commercials, we can measure similarities and identify a target video from thousands of videos. To process all the TV programs of 24 hours a day, feature extraction and matching process must be done in real-time. In this paper, we designed the visual feature DB for real-time processing and implemented real-time TV commercial monitoring system. To construct the DB, we extracted scene change information, block-based dominant colors and edge pattern histograms of TV commercial clips.