Exploring Motion and Appearance Information for Temporal Sentence Grounding
This work solves the problem of accurately grounding sentences in video sequences for applications like video understanding, though it is incremental by building on existing object detection methods.
The paper tackles temporal sentence grounding by addressing the limitations of previous methods that fail to distinguish ambiguous video frames due to frame-level features and lack motion analysis in object-level features, proposing a Motion-Appearance Reasoning Network (MARN) that incorporates both motion-aware and appearance-aware object features, resulting in significant performance improvements on Charades-STA and TACoS datasets.
This paper addresses temporal sentence grounding. Previous works typically solve this task by learning frame-level video features and align them with the textual information. A major limitation of these works is that they fail to distinguish ambiguous video frames with subtle appearance differences due to frame-level feature extraction. Recently, a few methods adopt Faster R-CNN to extract detailed object features in each frame to differentiate the fine-grained appearance similarities. However, the object-level features extracted by Faster R-CNN suffer from missing motion analysis since the object detection model lacks temporal modeling. To solve this issue, we propose a novel Motion-Appearance Reasoning Network (MARN), which incorporates both motion-aware and appearance-aware object features to better reason object relations for modeling the activity among successive frames. Specifically, we first introduce two individual video encoders to embed the video into corresponding motion-oriented and appearance-aspect object representations. Then, we develop separate motion and appearance branches to learn motion-guided and appearance-guided object relations, respectively. At last, both motion and appearance information from two branches are associated to generate more representative features for final grounding. Extensive experiments on two challenging datasets (Charades-STA and TACoS) show that our proposed MARN significantly outperforms previous state-of-the-art methods by a large margin.