In one embodiment, a video stream is received, and a spatially distinct region of a frame in the received video stream to be modified is identified, the spatially distinct region of the frame being encoded separately from any other region in the frame. A segment of the spatially distinct region of the frame to be modified is extracted. The extracted segment of the spatially distinct region is modified. The extracted segment of the spatially distinct region into a single-segment spatially distinct region is encoded. A network abstraction layer (NAL) header is associated with the encoded single-segment spatially distinct region. The encoded single-segment spatially distinct region and its associated NAL header is inserted into the received video stream following the identified spatially distinct region of the frame to be modified. Related methods, systems and apparatus are also described.