University of California

S Shankar Sastry

Kannan Ramchandran

Roozbeh Jafari

Ruzena Bajcsy

Chuohao Yeo

Allen Yang Yang

Parvez Ahammad

Methods for classifying aggregated data in a distributed sensor system are provided and illustrated with a wearable motion sensor network. The classification is operated in a distributed fashion on individual sensor nodes and a base station computer. The method classifies actions using a set of training motion sequences as prior examples and may reject outlying actions that are not in the training categories. Acquired sensor data is processed at the node by taking projections of the data to reduce dimensionality, calculating sparse representations of features using training sequences; validating and classifying local measurements and then transmitting classified measurements to a network base station. The base station aggregates local sensor measurements and performs a global classification of the data by forming global features from the local measurements; calculating sparse representations of global features; validating and classifying valid global features; and labeling global features and their corresponding local features.

System for detection of body motion

An apparatus for detecting an action in a test video. In an illustrative embodiment, the apparatus includes a first mechanism for receiving a query for a particular action via a query video. A second mechanism employs motion vectors associated with the test video to compute one or more motion-similarity values. The one or more motion-similarity values represent motion similarity between a first group of pixels in a first frame of a query video and a second group of pixels in a second frame of the test video based on the motion vectors. A third mechanism uses the one or more similarity values to search for the particular action or similar action in the test video. In a more specific embodiment, another mechanism aggregates the similarity values over a predetermined number of frames to facilitate estimating where the particular action or version thereof occurs or is likely to occur in the test video.

High speed video action recognition and localization

Growing community of inventors

Berkeley, CA, United States of America

S Shankar Sastry