Unifying Event Detection and Captioning as Sequence Generation via Pre-Training
[Paper]
The code has been tested on PyTorch 1.7.1.

Prerequisites

• Python 3.6
• Pytorch >= 1.6.0

Usage

• For ANET,

  sh driver/run.sh