COMPUTER VISION AND PATTERN RECOGNITION TASKS VIA PLAYSOURCING
Pattern recognition is used for a variety of applications in different techonogical fields. Within medical science, pattern recognition is the basis for computer-aided diagnosis (CAD) systems. CAD describes a procedure that supports the doctor‘s interpretations and findings. Other typical applications of pattern recognition techniques are automatic speech recognition, classification of text into several categories (e.g., spam/non-spam email messages), the automatic recognition of handwritten postal codes on postal envelopes or automatic recognition of images of human faces.1 A large quantity of algotithms has been developed for pattern recognition tasks. The invention describes a new crowdsourcing system for solving pattern recognition and computer vision tasks.
Being considered as a valid solution to the lack of ground truth data problem, crowdsourcing has recently gained a lot of attention within the biomedical domain. However, available concepts in life science domain require expert knowledge and thereby restrict the access to only very specific communities. The invention includes a novel concept for seamlessly embedding biomedical science into a common game canvas. In this way the requirements for prior knowledge are eliminated.2 The new technology relates to a system for the solution of computer vision and pattern recognition tasks via gamification, i.e., the usage of game concepts and technology in non-game contexts, and playsourcing. Playsourcing means the special form of crowdsourcing as the process of soliciting contributions from an online community of game players. The invention comprises of a general system for integrating visual and cognitive tasks into computer games such that all player-generated data can then be utilized to solve classical computer vision and pattern recognition tasks. Such tasks include but are not restricted to edge detection, detection of correspondences, image segmentation, object detection, classification of textures, objects, scenes, etc., optical flow estimation, depth estimation, 3D reconstruction, etc. Thereby, the described system represents a means for the integration of such tasks into existing computer games or at least combining these tasks with existing game concepts.
- Tumor segmentation
- Processing of hostology data
- Processing of image data
Proof of concept
Organ Segmentation via Gamification: After resampling the original image data with respect to a polar grid (left), the resampled data form the canvas for an arcade game (e.g. Floppy Bird) (middle). With one task of the game being separation of the horizon via the flight level the optimal segmentation (right) is characterized by the optimal fulfillment of the task.
 Milewski, Robert; Govindaraju, Venu (31 March 2008). „Binarization and cleanup of handwritten text from carbon copy medical form images“. Pattern
Recognition. 41 (4): 1308–1315.