• Database website online!
  • MAD database

    The MAD database was recorded using a Microsoft Kinect sensor in an indoor environment. The database contains:

  • The multimodal activity of 20 subjects recoded with the Microsoft Kinect camera. The modalities include: RGB video (240*320), 3D depth (240*320), and skeleton (3D coordinates of 20 joints per frame).
  • The activity of 20 subjects performing 35 sequential actions (see actionlist). Each subject repeats the set of 35 actions twice. Each video is about (4000-7000 frames) (See actionlist).
  • Ground truth labels: start and end of the action has been labeled, suitable for detection (also for classification).
  • Citation

    Dong Huang, Yi Wang*, Shitong Yao* and F. De la Torre. Sequential Max-Margin Event Detectors, ECCV 2014


    This work was partially funded by Samsung Electronics and partially supported by the National Science Foundation (NSF) under the grant RI-1116583. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the NSF.