Disney Theme Park Researchers Unveil Face-Tracking Video Surveillance Software

Disney theme park researchers reveal a new software that can follow the same subject across a network of surveillance cameras.

Steve Karantzoulidis Leave a Comment
Disney Theme Park Researchers Unveil Face-Tracking Video Surveillance Software

The Disney theme park “imagineers” have designed a new video surveillance software capable of tracking faces around the park. The software relies on a ‘deep spatial pyramid’ method for person re-identification.

Here is the abstract from their research paper:

Re-identification refers to the task of finding the same subject across a network of surveillance cameras. This task must deal with appearance changes caused by variations in illumination, a person’s pose, camera viewing angle and background clutter. State-of-the-art approaches usually focus either on feature modeling – designing image descriptors that are robust to changes in imaging conditions, or dissimilarity functions – learning effective metrics to compare images from different cameras.

Typically, with novel deep architectures both approaches can be merged into a single end-to-end training, but to become effective, this requires annotating thousands of subjects in each camera pair. Unlike standard CNN-based approaches, we introduce a spatial pyramid-like structure to the image and learn CNNs for image sub-regions at different scales. When training a CNN using only image sub-regions, we force the model to recognize not only the person’s identity but also the spatial location of the sub-region. This results in highly effective feature representations, which when combined with Mahalanobis-like metric learning significantly outperform state-of-the-art approaches.

Read next: Disney World’s Newest Technology Will Bring Out Your Inner Peter Pan

To put it simply: instead of identifying the full face of a person, the Disney theme park video surveillance software only identifies small rectangular sub-regions (patches) of a person’s face.

This allows for easier identification and tracking in poor lighting and background conditions and different camera angles due to the model recognizing not only the person’s identity, but also the spatial location of the sub-region.

After experimenting with the software, the research team found that the deep spatial pyramid software was more effective than the previously used detection systems. While park-goers may appreciate Disney’s focus on improving security, there will no doubt be backlash over privacy concerns. The New York City Police Department was sued this past May for not releasing documents related to its use of facial recognition technology.

See the original article on our sister publication, Security Sales & Integration.