# ID-VTG: Image-Disambiguated Video Temporal Grounding

arXiv: https://arxiv.org/abs/2608.20127 · HTML: https://mlsanity.com/paper/2608.20127/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 50 references that are in the map.
- First posted on arXiv 20 Aug 2026.

## Builds on (most cited first, 22 of 22)

- [Dense-Captioning Events in Videos](https://mlsanity.com/paper/1705.00754/) (2017)
- [Localizing Moments in Video with Natural Language](https://mlsanity.com/paper/1708.01641/) (2017)
- [TALL: Temporal Activity Localization via Language Query](https://mlsanity.com/paper/1705.02101/) (2017)
- [Learning 2D Temporal Adjacent Networks for Moment Localization with Natural Language](https://mlsanity.com/paper/1912.03590/) (2019)
- [FineGym: A Hierarchical Video Dataset for Fine-grained Action Understanding](https://mlsanity.com/paper/2004.06704/) (2020)
- [Local-Global Video-Text Interactions for Temporal Grounding](https://mlsanity.com/paper/2004.07514/) (2020)
- [Where Does It Exist: Spatio-Temporal Video Grounding for Multi-Form Sentences](https://mlsanity.com/paper/2001.06891/) (2020)
- [Negative Sample Matters: A Renaissance of Metric Learning for Temporal Grounding](https://mlsanity.com/paper/2109.04872/) (2021)
- [Human-centric Spatio-Temporal Video Grounding With Visual Transformers](https://mlsanity.com/paper/2011.05049/) (2020)
- [VLG-Net: Video-Language Graph Matching Network for Video Grounding](https://mlsanity.com/paper/2011.10132/) (2020)
- [SnAG: Scalable and Accurate Video Grounding](https://mlsanity.com/paper/2404.02257/) (2024)
- [Correlation-Guided Query-Dependency Calibration for Video Temporal Grounding](https://mlsanity.com/paper/2311.08835/) (2023)
- [Towards Balanced Alignment: Modal-Enhanced Semantic Modeling for Video Moment Retrieval](https://mlsanity.com/paper/2312.12155/) (2023)
- [Video Activity Localisation with Uncertainties in Temporal Boundary](https://mlsanity.com/paper/2206.12923/) (2022)
- [Advancing 3D Scene Understanding with MV-ScanQA Multi-View Reasoning Evaluation and TripAlign Pre-training Dataset](https://mlsanity.com/paper/2508.11058/) (2025)
- [MINOTAUR: Multi-task Video Grounding From Multimodal Queries](https://mlsanity.com/paper/2302.08063/) (2023)
- [Localizing Events in Videos with Multimodal Queries](https://mlsanity.com/paper/2406.10079/) (2024)
- [Hierarchical Event Memory for Accurate and Low-latency Online Video Temporal Grounding](https://mlsanity.com/paper/2508.04546/) (2025)
- [AR-VRM: Imitating Human Motions for Visual Robot Manipulation with Analogical Reasoning](https://mlsanity.com/paper/2508.07626/) (2025)
- [OmniGround: A Comprehensive Spatio-Temporal Grounding Benchmark for Real-World Complex Scenarios](https://mlsanity.com/paper/2511.16937/) (2025)
- [OmniVTG: A Large-Scale Dataset and Training Paradigm for Open-World Video Temporal Grounding](https://mlsanity.com/paper/2604.25276/) (2026)
- [Temporal-Aware Reasoning Optimization for Video Temporal Grounding](https://mlsanity.com/paper/2606.09248/) (2026)

## Abstract (arXiv)

Video Temporal Grounding (VTG) faces significant challenges when natural language queries must distinguish between multiple events involving visually similar entities, particularly when relying on fine-grained visual attributes that are difficult to describe accurately in words alone. To address this, we introduce Image-Disambiguated Video Temporal Grounding (ID-VTG), a task that leverages multimodal queries combining a reference image and a text description to precisely localize segments where a specific instance performs a described action. To facilitate research, we construct two benchmarks: IDVTG-Gym, focusing on fine-grained, compositionally ordered gymnastics actions with athletes in similar uniforms; and IDVTG-InternVid, an open-world dataset featuring diverse entities (e.g., humans, animals, fictional characters) and significant temporal distractors. Methodologically, we propose the Visually-Guided Disambiguation Aggregation (VGD-Agg) framework based on a dual-branch fast-slow architecture. The fast branch efficiently generates preliminary event proposals, while the slow branch performs fine-grained frame-level matching between video frames and the reference image. We enhance discriminability via two learnable tokens: a Compare Token, which represents hard negatives to probe for the presence of the target instance (as referred to by the query image), and a Depress Value, which represents text-irrelevant events. Proposals that the Compare Token identifies as lacking the target instance are pushed toward the Depress Value, thus easing disambiguation via the text query. Extensive experiments validate our approach, which achieves state-of-the-art results on the proposed benchmarks. Code is available at https://github.com/oceanflowlab/ID-VTG.