Apple researchers released LVSum, a human-annotated benchmark for testing how well AI models summarize long videos with accurate timestamps.
Continue to AI University →