Observers usually cannot discriminate the relative depth of a crowded feature with respect to crowding features about 2 arc min distant if all the features have the same luminance. However, stereo thresholds significantly less than 20 arc sec are obtained when the crowded feature is about twice as luminant as the features crowding it. The thresholds depend only upon the ratio of the luminance o...

This paper proposes Motion Structure Tracker (MST) to solve the problem of tracking in very crowded structured scenes. It combines visual tracking, motion pattern learning and multi-target tracking. Tracking in crowded scenes is very challenging due to hundreds of similar objects, cluttered background, small object size, and occlusions. However, structured crowded scenes exhibit clear motion pa...

Crowdedness caused by overlapping among similar objects is a ubiquitous challenge in the field of 2D visual object detection. In this paper, we first underline two main effects crowdedness issue: 1) IoU-confidence correlation disturbances (ICD) and 2) confused de-duplication (CDD). Then explore pathway cracking these nuts from perspective data augmentation. Primarily, particular copy- paste sch...

We consider the problem of data association in a multiperson tracking context. In semi-crowded environments, people are still discernible as individually moving entities, that undergo many interactions with other people in their direct surrounding. Finding the correct association is therefore difficult, but higher-order social factors, such as group membership, are expected to ease the problem....