Search for a command to run...
Benchmarking Vision-Language Models on Optical Character Recognition in Dynamic Video Environments