Usually Image recognition models are the ones used for video recognition too. Yolo models are often used for video recognition thanks to their near real-time inference time.

I invite you to check YoloV7 github page, they also have a script implementation of their model for video recognition on the main page.