Name
Abstract | ||
|---|---|---|
Many attempts have been made by researchers and scholars to make people feel more conversant to robots. One such example is the dance performance of an Entertainment Robot. In most cases, the challenge to program dance motions for a robot and synchronize them has been too heavy. In addition, pre-programmed dance moves and synchronization information are useful only for a specific music track and are useless for any other. To solve these problems, we developed a new system that can make a robot learn dance moves according to the input music track. The system comprises of two main parts: the first is a beat extraction system for music track; and the second one is a system that learns dance motion for Baxter. In the first part, music track is analyzed using STFT and peak-to-peak time duration is computed. This gives the beats per minute (BPM) of the given music track. The second part takes the BPM and duration of track and feeds it to the developed Q-learning algorithm to make Baxter learn dance moves and synchronize dance motion to beat rate. |
Year | DOI | Venue |
|---|---|---|
2015 | 10.1109/SYSOSE.2015.7151920 | 2015 10th System of Systems Engineering Conference (SoSE) |
Keywords | Field | DocType |
Q Learning,Learn dance,beat extraction,Robot Dance,Baxter | Computer vision,Beats per minute,Dance,Synchronization,Computer science,Q-learning,Real-time computing,Entertainment robot,Artificial intelligence,Beat (music),Robot | Conference |
Citations | PageRank | References |
0 | 0.34 | 13 |
Authors | ||
2 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Sulabh Kumra | 1 | 0 | 2.37 |
| Ferat Sahin | 2 | 706 | 45.49 |