# SGWIB:Sliced Gromov-Wasserstein Information Bottleneck for Video Highlight Detection

arXiv: https://arxiv.org/abs/2609.13966 · HTML: https://mlsanity.com/paper/2609.13966/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 27 of its 48 references that are in the map.
- First posted on arXiv 12 Sep 2026.

## Builds on (most cited first, 27 of 27)

- [Deep Variational Information Bottleneck](https://mlsanity.com/paper/1612.00410/) (2016)
- [MINE: Mutual Information Neural Estimation](https://mlsanity.com/paper/1801.04062/) (2018)
- [YouTube-8M: A Large-Scale Video Classification Benchmark](https://mlsanity.com/paper/1609.08675/) (2016)
- [CLUB: A Contrastive Log-ratio Upper Bound of Mutual Information](https://mlsanity.com/paper/2006.12013/) (2020)
- [Causal Attention for Unbiased Visual Recognition](https://mlsanity.com/paper/2108.08782/) (2021)
- [The Conditional Entropy Bottleneck](https://mlsanity.com/paper/2002.05379/) (2020)
- [Summarizing Videos with Attention](https://mlsanity.com/paper/1812.01969/) (2018)
- [Wasserstein Dependency Measure for Representation Learning](https://mlsanity.com/paper/1903.11780/) (2019)
- [Sliced Gromov-Wasserstein](https://mlsanity.com/paper/1905.10124/) (2019)
- [Rethinking the Evaluation of Video Summaries](https://mlsanity.com/paper/1903.11328/) (2019)
- [TR-DETR: Task-Reciprocal Transformer for Joint Moment Retrieval and Highlight Detection](https://mlsanity.com/paper/2401.02309/) (2024)
- [Cross-category Video Highlight Detection via Set-based Learning](https://mlsanity.com/paper/2108.11770/) (2021)
- [TimeX++: Learning Time-Series Explanations with Information Bottleneck](https://mlsanity.com/paper/2405.09308/) (2024)
- [CSTA: CNN-based Spatiotemporal Attention for Video Summarization](https://mlsanity.com/paper/2405.11905/) (2024)
- [TimeExpert: An Expert-Guided Video LLM for Video Temporal Grounding](https://mlsanity.com/paper/2508.01699/) (2025)
- [Aligning Multimodal Representations through an Information Bottleneck](https://mlsanity.com/paper/2506.04870/) (2025)
- [A Variational Information Bottleneck Based Method to Compress Sequential Networks for Human Action Recognition](https://mlsanity.com/paper/2010.01343/) (2020)
- [Watch Video, Catch Keyword: Context-aware Keyword Attention for Moment Retrieval and Highlight Detection](https://mlsanity.com/paper/2501.02504/) (2025)
- [Unleash the Potential of CLIP for Video Highlight Detection](https://mlsanity.com/paper/2404.01745/) (2024)
- [Unsupervised Video Highlight Detection by Learning from Audio and Visual Recurrence](https://mlsanity.com/paper/2407.13933/) (2024)
- [Revisiting Counterfactual Regression through the Lens of Gromov-Wasserstein Information Bottleneck](https://mlsanity.com/paper/2405.15505/) (2024)
- [SummDiff: Generative Modeling of Video Summarization with Diffusion](https://mlsanity.com/paper/2510.08458/) (2025)
- [CVA: Context-aware Video-text Alignment for Video Temporal Grounding](https://mlsanity.com/paper/2603.24934/) (2026)
- [HierSum: A Global and Local Attention Mechanism for Video Summarization](https://mlsanity.com/paper/2504.18689/) (2025)
- [Test-Time Adaptation for Video Highlight Detection Using Meta-Auxiliary Learning and Cross-Modality Hallucinations](https://mlsanity.com/paper/2508.04924/) (2025)
- [AHA -- Predicting What Matters Next: Online Highlight Detection Without Looking Ahead](https://mlsanity.com/paper/2509.16421/) (2025)
- [TripleSumm: Adaptive Triple-Modality Fusion for Video Summarization](https://mlsanity.com/paper/2603.01169/) (2026)

## Abstract (arXiv)

Video highlight detection aims to identify temporally important segments that capture the most informative or engaging events in a video. Reliable prediction therefore requires not only discriminative segment representations but also preservation of the temporal relationships among neighboring and distant segments. The information bottleneck principle has proven effective for learning compact and task-relevant representations, yet it has not been explored for video highlight detection, and applying conventional formulations directly would overlook inter-segment relational structure and distort highlight relevant temporal organization during compression. We therefore introduce the Sliced Gromov-Monge Gap (SGMG), a structure aware regularizer that measures the excess relational distortion induced by a prescribed source-to-bottleneck mapping relative to an optimal sliced structural correspondence. Building on SGMG, we develop SGWIB, an information-bottleneck framework for single-modal video highlight detection that learns compact bottleneck representations while preserving inter-segment temporal structure. We further introduce Home-Away-Related Contextual Pseudo-Labels and a contextual disentanglement module that reduce sports-specific contextual bias by separating highlight oriented information from contextual patterns. Experiments on MrHiSum and MoSu show that SGWIB attains the best Kendall's tau, Spearman's rho, mAP@50, and mAP@30 among the compared single-modal methods on both datasets. On MrHiSum, the visual model improves the strongest previous results by 0.031, 0.031, 0.87, and 0.75 on these four metrics, respectively. These results show that structure-aware information-bottleneck regularization combined with contextual disentanglement improves segment-level highlight prediction.