Towards Fine-Grained Video Question Answering
This work addresses fine-grained video understanding for AI researchers, but it is incremental as it builds on existing video-language models with new components.
The paper tackles the problem of limited temporal and spatial granularity in Video Question Answering by introducing the MOMA-QA dataset and the SGVLM model, achieving superior performance and setting new benchmarks on VideoQA tasks.
In the rapidly evolving domain of video understanding, Video Question Answering (VideoQA) remains a focal point. However, existing datasets exhibit gaps in temporal and spatial granularity, which consequently limits the capabilities of existing VideoQA methods. This paper introduces the Multi-Object Multi-Actor Question Answering (MOMA-QA) dataset, which is designed to address these shortcomings by emphasizing temporal localization, spatial relationship reasoning, and entity-centric queries. With ground truth scene graphs and temporal interval annotations, MOMA-QA is ideal for developing models for fine-grained video understanding. Furthermore, we present a novel video-language model, SGVLM, which incorporates a scene graph predictor, an efficient frame retriever, and a pre-trained large language model for temporal localization and fine-grained relationship understanding. Evaluations on MOMA-QA and other public datasets demonstrate the superior performance of our model, setting new benchmarks for VideoQA.