mlsanityData to 7 Oct 2026

HieraMamba: Video Temporal Grounding via Hierarchical Anchor-Mamba Pooling

arXiv 2510.23043 ↗First posted 27 Oct 2025cs.CVcited by 1

On the map: State-space sequence models (S4, Mamba) → Video, motion and generative Mamba models

Abstract (arXiv)

Video temporal grounding, the task of localizing the start and end times of a natural language query in untrimmed video, requires capturing both global context and fine-grained temporal detail. This challenge is particularly pronounced in long videos, where existing methods often compromise temporal fidelity by over-downsampling or relying on fixed windows. We present HieraMamba, a hierarchical architecture that preserves temporal structure and semantic richness across scales. At its core are Anchor-MambaPooling (AMP) blocks, which utilize Mamba's selective scanning to produce compact anchor tokens that summarize video content at multiple granularities. Two complementary objectives, anchor-conditioned and segment-pooled contrastive losses, encourage anchors to retain local detail while remaining globally discriminative. HieraMamba sets a new state-of-the-art on Ego4D-NLQ, MAD, and TACoS, demonstrating precise, temporally faithful localization in long, untrimmed videos.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Placed in Video, motion and generative Mamba models, one of 10 directions in State-space sequence models (S4, Mamba). Builds on 69 of its 81 references that are in the map. First posted on arXiv 27 Oct 2025.

Builds on

69 of 81 references.

Closest ancestors · 5

Same direction, newest first

  1. 2025Vamba: Understanding Hour-Long Videos with Hybrid Mamba-Transformers14 Mar
  2. 2024VideoMamba: Spatio-Temporal Selective State Space Model11 Jul
  3. 2024Snakes and Ladders: Two Steps Up for VideoMamba27 Jun
  4. 2024Simba: Mamba augmented U-ShiftGCN for Skeletal Action Recognition in Videos11 Apr
  5. 2024VideoMamba: State Space Model for Efficient Video Understanding11 Mar

Other references · 64

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  2. 2023Mamba: Linear-Time Sequence Modeling with Selective State Spacescited by 4,616
  3. 2014Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translationcited by 4,016
  4. 2021Efficiently Modeling Long Sequences with Structured State Spacescited by 2,424
  5. 2020GLU Variants Improve Transformercited by 1,724
  6. 2021Is Space-Time Attention All You Need for Video Understanding?cited by 1,609
  7. 2021Ego4D: Around the World in 3,000 Hours of Egocentric Videocited by 1,530
  8. 2019Root Mean Square Layer Normalizationcited by 1,425
  9. 2024Transformers are SSMs: Generalized Models and Efficient Algorithms Through Structured State Space Dualitycited by 1,283
  10. 2024VMamba: Visual State Space Modelcited by 955
  11. 2016Language Modeling with Gated Convolutional Networkscited by 688
  12. 2017Localizing Moments in Video with Natural Languagecited by 660
38 more
  1. 2021Combining Recurrent, Convolutional, and Continuous-time Models with Linear State-Space Layerscited by 607
  2. 2020HiPPO: Recurrent Memory with Optimal Polynomial Projectionscited by 600
  3. 2017TALL: Temporal Activity Localization via Language Querycited by 581
  4. 2019Distance-IoU Loss: Faster and Better Learning for Bounding Box Regressioncited by 454
  5. 2022InternVideo: General Video Foundation Models via Generative and Discriminative Learningcited by 441
  6. 2024Jamba: A Hybrid Transformer-Mamba Language Modelcited by 390
  7. 2017Soft-NMS -- Improving Object Detection With One Line of Codecited by 380
  8. 2022ActionFormer: Localizing Moments of Actions with Transformerscited by 308
  9. 2019Learning 2D Temporal Adjacent Networks for Moment Localization with Natural Languagecited by 283
  10. 2019G-TAD: Sub-Graph Localization for Temporal Action Detectioncited by 248
  11. 2024VideoTree: Adaptive Tree-based Video Representation for LLM Reasoning on Long Videoscited by 242
  12. 2020Span-based Localizing Network for Natural Language Video Localizationcited by 234
  13. 2022Egocentric Video-Language Pretrainingcited by 231
  14. 2024MambaVision: A Hybrid Mamba-Transformer Vision Backbonecited by 222
  15. 2019Graph Convolutional Networks for Temporal Action Localizationcited by 208
  16. 2023UniVTG: Towards Unified Video-Language Temporal Groundingcited by 189
  17. 2020Local-Global Video-Text Interactions for Temporal Groundingcited by 188
  18. 2020Dense Regression Network for Video Groundingcited by 176
  19. 2019Semantic Conditioned Dynamic Modulation for Temporal Sentence Grounding in Videoscited by 135
  20. 2021Context-aware Biaffine Localizing Network for Temporal Sentence Groundingcited by 107
  21. 2021MAD: A Scalable Dataset for Language Grounding in Videos from Movie Audio Descriptionscited by 107
  22. 2022Long Movie Clip Classification with State-Space Video Modelscited by 106
  23. 2021Negative Sample Matters: A Renaissance of Metric Learning for Temporal Groundingcited by 90
  24. 2024Video ReCap: Recursive Captioning of Hour-Long Videoscited by 88
  25. 2019ExCL: Extractive Clip Localization Using Natural Language Descriptionscited by 82
  26. 2022TALLFormer: Temporal Action Localization with a Long-memory Transformercited by 78
  27. 2025Dynamic Chunking for End-to-End Hierarchical Sequence Modelingcited by 65
  28. 2022Efficient Movie Scene Detection using State-Space Transformerscited by 51
  29. 2020VLG-Net: Video-Language Graph Matching Network for Video Groundingcited by 50
  30. 2023AutoAD: Movie Description in Contextcited by 50
  31. 2022Memory-Guided Semantic Learning Network for Temporal Sentence Groundingcited by 46
  32. 2023NaQ: Leveraging Narrations as Queries to Supervise Episodic Memorycited by 46
  33. 2025STORM: Token-Efficient Long Video Understanding for Multimodal LLMscited by 46
  34. 2024Hydra: Bidirectional State Space Models Through Generalized Matrix Mixerscited by 43
  35. 2021Adaptive Proposal Generation Network for Temporal Sentence Localization in Videoscited by 42
  36. 2024SnAG: Scalable and Accurate Video Groundingcited by 36
  37. 2021Progressively Guide to Attend: An Iterative Alignment Framework for Temporal Sentence Groundingcited by 31
  38. 2023Helping Hands: An Object-Aware Ego-Centric Video Recognition Modelcited by 31

50 most cited of 64 All 64 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026UniversalVTG: A Universal and Lightweight Foundation Model for Video Temporal Groundingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Universal Video Temporal Grounding with Generative Multi-modal Large Language Models
  2. 2026DART: Difficulty-Adaptive Routing for Zero-Shot Video Temporal Grounding
  3. 2025MLVTG: Mamba-Based Feature Alignment and LLM-Driven Purification for Multi-Modal Video Temporal Grounding
  4. 2026Conditional Multi-Event Temporal Grounding in Long-Form Video
  5. 2026Towards One-to-Many Temporal Grounding
  6. 2024Training-free Video Temporal Grounding using Large-scale Pre-trained Models
  7. 2025Enrich and Detect: Video Temporal Grounding with Multimodal LLMs
  8. 2026OmniVTG: A Large-Scale Dataset and Training Paradigm for Open-World Video Temporal Grounding