Object Relational Graph with Teacher-Recommended Learning for Video Captioning

TL;DR

This paper introduces an object relational graph encoder and a teacher-recommended learning strategy that leverage external language models to improve video captioning by enriching visual features and addressing long-tailed word distributions.

Contribution

The paper presents a novel object relational graph encoder and a teacher-recommended learning method that effectively incorporate external language knowledge into video captioning models.

Findings

Achieves state-of-the-art results on MSVD, MSR-VTT, and VATEX datasets.

Enriches visual representation with object interaction features.

Improves handling of long-tailed word distributions in captioning.

Abstract

Taking full advantage of the information from both vision and language is critical for the video captioning task. Existing models lack adequate visual representation due to the neglect of interaction between object, and sufficient training for content-related words due to long-tailed problems. In this paper, we propose a complete video captioning system including both a novel model and an effective training strategy. Specifically, we propose an object relational graph (ORG) based encoder, which captures more detailed interaction features to enrich visual representation. Meanwhile, we design a teacher-recommended learning (TRL) method to make full use of the successful external language model (ELM) to integrate the abundant linguistic knowledge into the caption model. The ELM generates more semantically similar word proposals which extend the ground-truth words used for training to deal with the long-tailed problem. Experimental evaluations on three benchmarks: MSVD, MSR-VTT and VATEX show the proposed ORG-TRL system achieves state-of-the-art performance. Extensive ablation studies and visualizations illustrate the effectiveness of our system.