Spatio-Temporal Attention Models for Grounded Video Captioning

TL;DR

This paper introduces a spatio-temporal attention model for grounded video captioning that localizes visual concepts over space and time without grounding supervision, achieving state-of-the-art results.

Contribution

It proposes a novel deep neural network combining spatio-temporal attention with image classification for improved video captioning and localization.

Findings

Achieves state-of-the-art results on YouTube captioning benchmark.

Effectively localizes visual concepts without grounding supervision.

Enhances content localization and generalization in video captioning.

Abstract

Automatic video captioning is challenging due to the complex interactions in dynamic real scenes. A comprehensive system would ultimately localize and track the objects, actions and interactions present in a video and generate a description that relies on temporal localization in order to ground the visual concepts. However, most existing automatic video captioning systems map from raw video data to high level textual description, bypassing localization and recognition, thus discarding potentially valuable information for content localization and generalization. In this work we present an automatic video captioning model that combines spatio-temporal attention and image classification by means of deep neural network structures based on long short-term memory. The resulting system is demonstrated to produce state-of-the-art results in the standard YouTube captioning benchmark while also offering the advantage of localizing the visual concepts (subjects, verbs, objects), with no grounding supervision, over space and time.