Spatio-Temporal Attention Models for Grounded Video Captioning
This work addresses the challenge of generating grounded descriptions for videos, which is important for applications like accessibility and video analysis, though it builds incrementally on existing attention-based methods.
The authors tackled the problem of automatic video captioning by developing a model that combines spatio-temporal attention and image classification using deep neural networks, achieving state-of-the-art results on the YouTube captioning benchmark and enabling localization of visual concepts without supervision.
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.