How to make a 3d movie?

Has anybody tried to make a synthetic 3D movie dataset for training object recognition algorithms?

• I plan to experiment with a scriptable 3D & physics engine such as Panda3D to generate movies of 3D scenes from simple geometric moving objects (e.g. cubes, spheres, ...) to real-life looking scenes (e.g. city streets with buildings, animated cars and pedestrians) so as to prepare semantically annotated datasets for computer vision tasks and try to experiment with curriculum learning for deep semi-supervised learning algorithms. Is there any existing implementation of this approach?

• Answer:

  Yes. See the work of , James DiCarlo, and David Cox, e.g. http://pinto.scripts.mit.edu/uploads/Research/pinto-dicarlo-cox-eccv-2008-lfw_final.pdf, http://pinto.scripts.mit.edu/uploads/Research/pinto-cox-dicarlo-plos-2008_final.pdf, http://pinto.scripts.mit.edu/uploads/Research/pinto-dicarlo-cox-cvpr-2009-mkl.pdf, and http://pinto.scripts.mit.edu/uploads/Research/soainv.pdf. This group uses artificial visual datasets to test the ability of learning methods to handle variation and to develop invariant feature representations.