In-sensor processing of dynamic and static information of visual objects avoids exchanging redundant data between physically separated sensing and computing units, holding promise for computer vision ...