Unsupervised Video Scoring system for Behavioral research experiments
Behavioral assessment of mutant mouse models and novel candidate drugs is a slow and labor intensive process. This limitation produces a signiifcant delay in the development cycle of a new dgur discovery. Behavioral tests are a great tool for preclinical studies on how a new drug could or may affect but they are slow and labor-intensive. One of the greatest delay factors in the need of continuous monitoring of mouse models and the subsequent human scoring. Except the obvious time limits from the involvement of human scorers the factor of subjectivity is another issue that could bias the scoring results. Towards to the direction of design and develop an unbiased, fully automated system for Behavioral research experiments, TTSS unit proposes a video scoring system (named VisionLab) that can be operated fully unsupervised. Using simple but efficient Computer Vision algorithms, VisionLab is able to complete succesfully (in comparison with experienced human scorers) the scoring procedure.