Based on the Qwen-VL large-scale vision-language model, it automatically recognizes and processes videos. By defining a descriptive parameter, it presents each parametric processing result over tim…