Mirror-Fusion Attention for Reflection-Aware Self-Supervised Representation Learning

• URL: http://arxiv.org/abs/2607.00850v1
• Date: Wed, 01 Jul 2026 12:16:36 GMT
• Title: Mirror-Fusion Attention for Reflection-Aware Self-Supervised Representation Learning
• Abstract summary: Mirror-Fusion-Augmented Self-Supervised Learning (MFASSL) is a Vision Transformer framework that injects a soft reflection prior into standard SSL.<n>MFASSL constructs mirror-paired views aligned to an estimated symmetry axis and introduces a lightweight Mirror-Fusion Attention (MFA) module.
• Score: 5.016334858905707
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Most self-supervised learning (SSL) methods encourage invariance across augmentations, but strict flip invariance can suppress informative left--right correspondences in approximately bilateral data such as medical images and human faces. We propose Mirror-Fusion-Augmented Self-Supervised Learning (MFASSL), a Vision Transformer framework that injects a soft reflection prior into standard SSL without redesigning the backbone. MFASSL constructs mirror-paired views aligned to an estimated symmetry axis and introduces a lightweight Mirror-Fusion Attention (MFA) module for adaptive token-level interaction between mirrored regions while preserving asymmetric cues. The base SSL objective is further coupled with reflection-consistency and mid-layer token-alignment losses. Across CheXpert, BraTS, CelebA-HQ, and WFLW, MFASSL improves downstream performance, calibration, and reflection robustness over MoCo-v3, DINO, and MAE baselines under matched ViT-B/16 settings. It also achieves stronger and more consistent gains than recent equivariant SSL approaches with only approximately 2.7\% additional parameters. These results show that lightweight geometry-aware priors can effectively complement invariance-based SSL.

Related papers

• Geometry-Aware Contrastive Learning for Few-Shot Automatic Modulation Recognition [5.310289270050359]
  Dynamic-Adaptive Contrastive Learning (DyCo-CL) is a geometry-aware framework that couples Virtual Adrialversa Augmentation (VAA) with a semantic consistency loss.<n>DyCo-CL achieves a 6.27% accuracy gain in 1-shot settings over prior methods.
  arXiv  Detail & Related papers  (2026-05-26T06:33:57Z)
• Reflex: Reinforcement Learning with Reflection Symmetry Exploitation in State-Based Continuous Control [10.639649058223114]
  We introduce Reflex, a paradigm that seamlessly integrates with both on-policy and off-policy RL algorithms.<n>Building on a theoretical analysis of symmetry-preserving optimal value functions and policies, Reflex integrates reflection symmetry into policy learning.<n>We integrate Reflex with PPO and SAC, and evaluate it on a suite of OpenAI Gym and DeepMind Control benchmarks.
  arXiv  Detail & Related papers  (2026-05-22T09:24:44Z)
• MirrorLA: Reflecting Feature Map for Vision Linear Attention [49.41670925034762]
  Linear attention significantly reduces the computational complexity of Transformers from quadratic to linear, yet it consistently lags behind softmax-based attention in performance.<n>We propose MirrorLA, a geometric framework that substitutes passive truncation with active reorientation.<n>MirrorLA achieves state-of-the-art performance across standard benchmarks, demonstrating that strictly linear efficiency can be achieved without compromising representational fidelity.
  arXiv  Detail & Related papers  (2026-02-04T09:14:09Z)
• Feature-Space Adversarial Robustness Certification for Multimodal Large Language Models [59.6491828112519]
  Multimodal large language models (MLLMs) exhibit strong capabilities across diverse applications.<n> MLLMs are vulnerable to adversarial perturbations that distort their feature representations and induce erroneous predictions.<n>We propose Feature-space Smoothing (FS), a general framework that provides certified robustness guarantees at the feature representation level of MLLMs.
  arXiv  Detail & Related papers  (2026-01-22T18:52:21Z)
• Enhancing CLIP Robustness via Cross-Modality Alignment [54.01929554563447]
  We propose Cross-modality Alignment, an optimal transport-based framework for vision-language models.<n> COLA restores global image-text alignment and local structural consistency in the feature space.<n> COLA is training-free and compatible with existing fine-tuned models.
  arXiv  Detail & Related papers  (2025-10-28T03:47:44Z)
• MergeMix: A Unified Augmentation Paradigm for Visual and Multi-Modal Understanding [23.96717124380285]
  MergeMix is a training-time augmentation paradigm that bridges SFT and RL.<n>It first applies an attention-aware image mixing via token merge with more cluster representation and spatial context.<n>It then presents a preference-driven training paradigm for MLLMs by building preference pairs with mixed images and raw images, and optimizing via SimPO loss.
  arXiv  Detail & Related papers  (2025-10-27T16:12:40Z)
• RL makes MLLMs see better than SFT [96.508432109136]
  We conduct a critical yet under-explored analysis of the vision encoder of Multimodal Language Model (MLLM)<n>Our results demonstrate that MLLM's post-training strategy (i.e., SFT or RL) not only leads to distinct outcomes on MLLM downstream tasks, but also fundamentally reshapes MLLM's underlying visual representations.<n>We then reframe our findings into a simple recipe for building strong vision encoders for MLLMs, Preference-Instructed Vision OpTimization (PIVOT)
  arXiv  Detail & Related papers  (2025-10-18T03:37:17Z)
• Cross-Modal Consistency Learning for Sign Language Recognition [92.44927164283641]
  Existing pre-training methods solely focus on the compact pose data.<n>We propose a Cross-modal Consistency Learning framework (CCL- SLR)<n>CCL- SLR learns from both RGB and pose modalities based on self-supervised pre-training.
  arXiv  Detail & Related papers  (2025-03-16T12:34:07Z)
• MAGMA: Manifold Regularization for MAEs [1.7478203318226315]
  Masked Autoencoders (MAEs) are an important divide in self-supervised learning (SSL)<n>We introduce MAGMA, a novel batch-wide layer-wise regularization loss applied to representations of different Transformer layers.<n>We demonstrate that by plugging in the proposed regularization loss, one can significantly improve the performance of MAE-based models.
  arXiv  Detail & Related papers  (2024-12-03T22:14:10Z)
• Symmetry-Aware Transformer-based Mirror Detection [85.47570468668955]
  We propose a dual-path Symmetry-Aware Transformer-based mirror detection Network (SATNet) SATNet includes two novel modules: Symmetry-Aware Attention Module (SAAM) and Contrast and Fusion Decoder Module (CFDM) Experimental results show that SATNet outperforms both RGB and RGB-D mirror detection methods on all available mirror detection datasets.
  arXiv  Detail & Related papers  (2022-07-13T16:40:01Z)
• Directional Self-supervised Learning for Risky Image Augmentations [54.43314754436954]
  We propose a directional self-supervised learning paradigm (DSSL) for robust augmentation policies. DSSL can be easily implemented with a few lines of Pseudocode and is highly flexible to popular self-supervised learning frameworks.
  arXiv  Detail & Related papers  (2021-10-26T10:33:25Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.