TQA (Temporal Question Answering)

Dataset of 30,000 video records for fine-tuning time-sensitive vision models.