Clear Roles and Previews for Two-Input Upload Forms

A photo-to-motion form requires a still image and a reference video, and the interface should make the purpose of each file obvious before generation begins. Labels such as "Character photo" and "Dance reference" explain what each input contributes, rather than generic "Upload file" boxes. Previews should sit beside each input, with the image showing the subject clearly and the video allowing its chosen movement to be inspected. Because long clips often contain only a short segment, the form should display the selected start and end points, let users play that passage, and update the summary when the range changes. Distinct status labels for uploading versus ready states, plus errors kept beside the relevant input, help users fix problems without re-entering the whole form.