Towards Parameter-Efficient Integration of Pre-Trained Language Models In Temporal Video Grounding

TL;DR

This paper investigates the integration of pre-trained language models into Temporal Video Grounding, demonstrating that NLP adapters enable efficient fine-tuning and improve model performance without altering visual inputs.

Contribution

It provides a systematic study of PLM effects in TVG and evaluates parameter-efficient adapters, highlighting their effectiveness and practical benefits.

Findings

PLMs significantly improve TVG performance without changing visual inputs.

NLP adapters can match state-of-the-art results with less computational cost.

Different adapters perform better in various scenarios.

Abstract

This paper explores the task of Temporal Video Grounding (TVG) where, given an untrimmed video and a natural language sentence query, the goal is to recognize and determine temporal boundaries of action instances in the video described by the query. Recent works tackled this task by improving query inputs with large pre-trained language models (PLM) at the cost of more expensive training. However, the effects of this integration are unclear, as these works also propose improvements in the visual inputs. Therefore, this paper studies the effects of PLMs in TVG and assesses the applicability of parameter-efficient training with NLP adapters. We couple popular PLMs with a selection of existing approaches and test different adapters to reduce the impact of the additional parameters. Our results on three challenging datasets show that, without changing the visual inputs, TVG models greatly benefited from the PLM integration and fine-tuning, stressing the importance of sentence query representation in this task. Furthermore, NLP adapters were an effective alternative to full fine-tuning, even though they were not tailored to our task, allowing PLM integration in larger TVG models and delivering results comparable to SOTA models. Finally, our results shed light on which adapters work best in different scenarios.