Our Semantic-based Cinemagraph Results

Input video

Ours (person)

Ours (water)

Ours (tree)