# Dense-Captioning Events in Videos

## TL;DR

This paper introduces the task of dense-captioning events in videos, proposing a new model capable of detecting and describing multiple events simultaneously, and presents a large-scale benchmark dataset for evaluation.

## Contribution

It presents a novel model that detects and describes multiple events in videos in a single pass, and introduces the ActivityNet Captions dataset for dense-captioning evaluation.

## Key findings

- The model effectively detects and describes multiple events in videos.
- The ActivityNet Captions dataset provides a large-scale benchmark for dense-captioning.
- Performance results demonstrate the model's capability in event detection and description.

## Abstract

Most natural videos contain numerous events. For example, in a video of a "man playing a piano", the video might also contain "another man dancing" or "a crowd clapping". We introduce the task of dense-captioning events, which involves both detecting and describing events in a video. We propose a new model that is able to identify all events in a single pass of the video while simultaneously describing the detected events with natural language. Our model introduces a variant of an existing proposal module that is designed to capture both short as well as long events that span minutes. To capture the dependencies between the events in a video, our model introduces a new captioning module that uses contextual information from past and future events to jointly describe all events. We also introduce ActivityNet Captions, a large-scale benchmark for dense-captioning events. ActivityNet Captions contains 20k videos amounting to 849 video hours with 100k total descriptions, each with it's unique start and end time. Finally, we report performances of our model for dense-captioning events, video retrieval and localization.