AR engines modify a picture or video of the real world to include more than simply real world content. The modified image/frame content is passed along to one or multiple downstream AR engines which treat the content as original input. Modifications to images/frames may include the addition of virtual markers. Virtual markers may be visual content of virtual origin that is added to an image or frame. A virtual marker may trigger a predetermined reaction from a downstream AR engine when that downstream AR engine processes the modified content. For example, a virtual marker added to an image by an upstream AR engine may trigger a downstream AR engine to output a particular augmentation. The virtual marker may have a known meaning to both the upstream AR engine and the downstream AR engine. Accordingly, there may be “collaboration” among AR engines and a reduction in the processing requirements of downstream and overall image processing.