mlsanityData to 7 Oct 2026

Famba-V: Fast Vision Mamba with Cross-Layer Token Fusion

arXiv 2409.09808 ↗First posted 15 Sep 2024cs.CV cs.AIcited by 8

On the map: State-space sequence models (S4, Mamba) → Vision Mamba backbones

Abstract (arXiv)

Mamba and Vision Mamba (Vim) models have shown their potential as an alternative to methods based on Transformer architecture. This work introduces Fast Mamba for Vision (Famba-V), a cross-layer token fusion technique to enhance the training efficiency of Vim models. The key idea of Famba-V is to identify and fuse similar tokens across different Vim layers based on a suit of cross-layer strategies instead of simply applying token fusion uniformly across all the layers that existing works propose. We evaluate the performance of Famba-V on CIFAR-100. Our results show that Famba-V is able to enhance the training efficiency of Vim models by reducing both training time and peak memory usage during training. Moreover, the proposed cross-layer strategies allow Famba-V to deliver superior accuracy-efficiency trade-offs. These results all together demonstrate Famba-V as a promising efficiency enhancement technique for Vim models.

Key facts. Cited by 8 papers in the map (data through 7 Oct 2026). Placed in Vision Mamba backbones, one of 10 directions in State-space sequence models (S4, Mamba). Builds on 32 of its 35 references that are in the map. First posted on arXiv 15 Sep 2024.

Builds on

32 of 35 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024MambaVC: Learned Visual Compression with Selective State Spaces24 May
  2. 2024DVMSR: Distillated Vision Mamba for Efficient Super-Resolution5 May
  3. 2024EfficientVMamba: Atrous Selective Scan for Light Weight Visual Mamba15 Mar
  4. 2024Vision Mamba: Efficient Visual Representation Learning with Bidirectional State Space Model17 Jan

Other references · 28

  1. 2017Attention Is All You Needcited by 59,932
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  3. 2023Mamba: Linear-Time Sequence Modeling with Selective State Spacescited by 4,616
  4. 2021Efficiently Modeling Long Sequences with Structured State Spacescited by 2,424
  5. 2024Transformers are SSMs: Generalized Models and Efficient Algorithms Through Structured State Space Dualitycited by 1,283
  6. 2022Token Merging: Your ViT But Fastercited by 995
  7. 2021DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsificationcited by 849
  8. 2021Combining Recurrent, Convolutional, and Continuous-time Models with Linear State-Space Layerscited by 607
  9. 2021AdaViT: Adaptive Tokens for Efficient Vision Transformercited by 360
  10. 2022Not All Patches are What You Need: Expediting Vision Transformers via Token Reorganizationscited by 359
  11. 2022Diagonal State Spaces are as Effective as Structured State Spacescited by 350
  12. 2022Long Range Language Modeling via Gated State Spacescited by 265
16 more
  1. 2024SVD-LLM: Truncation-aware Singular Value Decomposition for Large Language Model Compressioncited by 231
  2. 2021AdaViT: Adaptive Vision Transformers for Efficient Image Recognitioncited by 228
  3. 2023Efficient Large Language Models: A Surveycited by 183
  4. 2021PTQ4ViT: Post-training quantization for vision transformers with twin uniform quantizationcited by 173
  5. 2024Samba: Simple Hybrid State Space Models for Efficient Unlimited Context Language Modelingcited by 167
  6. 2021SPViT: Enabling Faster Vision Transformers via Soft Token Pruningcited by 165
  7. 2021Patch Slimming for Efficient Vision Transformerscited by 131
  8. 2024Scaling Laws with Vocabulary: Larger Models Deserve Larger Vocabulariescited by 113
  9. 2023Selective Structured State-Spaces for Long-Form Video Understandingcited by 105
  10. 2024LOOK-M: Look-Once Optimization in KV Cache for Efficient Multimodal Long-Context Inferencecited by 93
  11. 2021Vision Transformer Pruningcited by 77
  12. 2023Towards Accurate Post-Training Quantization for Vision Transformercited by 70
  13. 2021Global Vision Transformer Pruning with Hessian-Aware Saliencycited by 69
  14. 2022NoisyQuant: Noisy Bias-Enhanced Post-Training Activation Quantization for Vision Transformerscited by 59
  15. 2023PuMer: Pruning and Merging Tokens for Efficient Vision Language Modelscited by 53
  16. 2024SpectralMamba: Efficient Mamba for Hyperspectral Image Classificationcited by 36

Built on by

Cited by year

Citing papers per year: 2025: 8.

8 papers.

  1. 2025SVD-LLM V2: Optimizing Singular Value Truncation for Large Language Model Compressioncited by 63
  2. 2025MEDA: Dynamic KV Cache Allocation for Efficient Multimodal Long-Context Inferencecited by 31
  3. 2025A Survey on Mamba Architecture for Vision Applicationscited by 8
  4. 2025Mamba-Sea: A Mamba-based Framework with Global-to-Local Sequence Augmentation for Generalizable Medical Image Segmentationcited by 7
  5. 2025UAVD-Mamba: Deformable Token Fusion Vision Mamba for Multimodal UAV Detectioncited by 2
  6. 2025COXNet: Cross-Layer Fusion with Adaptive Alignment and Scale Integration for RGBT Tiny Object Detectioncited by 2
  7. 2025Fast Vision Mamba: Pooling Spatial Dimensions for Accelerated Processingcited by 0
  8. 2025Sequential Token Merging: Revisiting Hidden Statescited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025ViM-VQ: Efficient Post-Training Vector Quantization for Visual Mamba
  2. 2024Faster Vision Mamba is Rebuilt in Minutes via Merged Token Re-training
  3. 2024Stochastic Layer-Wise Shuffle for Improving Vision Mamba Training
  4. 2026ViM-Q: Scalable Algorithm-Hardware Co-Design for Vision Mamba Model Inference on FPGA
  5. 2025Dynamic Vision Mamba
  6. 2025Post-Training Quantization for Vision Mamba with k-Scaled Quantization and Reparameterization
  7. 2025Training-free Token Reduction for Vision Mamba
  8. 2024PTQ4VM: Post-Training Quantization for Visual Mamba