Benchmarking Unlearning for Vision Transformers

• URL: http://arxiv.org/abs/2602.20114v1
• Date: Mon, 23 Feb 2026 18:33:16 GMT
• Title: Benchmarking Unlearning for Vision Transformers
• Authors: Kairan Zhao, Iurie Luca, Peter Triantafillou,
• Abstract summary: This work is the first to benchmarking machine unlearning (MU) algorithm performance in different Vision Transformers (VTs) and at different capacities.<n>It characterizes how VTs training data relative to CNNs, and assesses the impact of different proxies on performance.<n>Overall, this work offers a benchmarking basis, enabling reproducible, fair, and comprehensive comparisons of existing (and future) MU algorithms on VTs.
• Score: 4.9193859756091145
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Research in machine unlearning (MU) has gained strong momentum: MU is now widely regarded as a critical capability for building safe and fair AI. In parallel, research into transformer architectures for computer vision tasks has been highly successful: Increasingly, Vision Transformers (VTs) emerge as strong alternatives to CNNs. Yet, MU research for vision tasks has largely centered on CNNs, not VTs. While benchmarking MU efforts have addressed LLMs, diffusion models, and CNNs, none exist for VTs. This work is the first to attempt this, benchmarking MU algorithm performance in different VT families (ViT and Swin-T) and at different capacities. The work employs (i) different datasets, selected to assess the impacts of dataset scale and complexity; (ii) different MU algorithms, selected to represent fundamentally different approaches for MU; and (iii) both single-shot and continual unlearning protocols. Additionally, it focuses on benchmarking MU algorithms that leverage training data memorization, since leveraging memorization has been recently discovered to significantly improve the performance of previously SOTA algorithms. En route, the work characterizes how VTs memorize training data relative to CNNs, and assesses the impact of different memorization proxies on performance. The benchmark uses unified evaluation metrics that capture two complementary notions of forget quality along with accuracy on unseen (test) data and on retained data. Overall, this work offers a benchmarking basis, enabling reproducible, fair, and comprehensive comparisons of existing (and future) MU algorithms on VTs. And, for the first time, it sheds light on how well existing algorithms work in VT settings, establishing a promising reference performance baseline.

Related papers

• TimeLens: Rethinking Video Temporal Grounding with Multimodal LLMs [81.78017865436816]
  We present TimeLens, a systematic investigation into building MLLMs with strong video temporal grounding ability.<n>We first expose critical quality issues in existing VTG benchmarks and introduce TimeLens-Bench, comprising meticulously re-annotated versions of three popular benchmarks.<n>We also address noisy training data through an automated re-annotation pipeline, yielding TimeLens-100K, a large-scale, high-quality training dataset.
  arXiv  Detail & Related papers  (2025-12-16T18:59:58Z)
• A Comparative Study of Vision Transformers and CNNs for Few-Shot Rigid Transformation and Fundamental Matrix Estimation [3.5684665108045377]
  Vision-transformers (ViTs) and large-scale convolution-neural-networks (CNNs) have reshaped computer vision through pretrained feature representations.<n>This work considers two such tasks: 1) estimating 2D rigid transformations between pairs of images and 2) predicting the fundamental matrix for stereo image pairs.<n> Empirical comparative analysis shows that, similar to training from scratch, ViTs outperform CNNs during refinement in large downstream-data scenarios.
  arXiv  Detail & Related papers  (2025-10-06T13:18:27Z)
• DyMU: Dynamic Merging and Virtual Unmerging for Efficient VLMs [124.52164183968145]
  We present DyMU, an efficient, training-free framework that reduces the computational burden of vision-language models (VLMs)<n>Our approach comprises two key components. First, Dynamic Token Merging (DToMe) reduces the number of visual token embeddings by merging similar tokens based on image complexity.<n>Second, Virtual Token Unmerging (VTU) simulates the expected token sequence for large language models (LLMs) by efficiently reconstructing the attention dynamics of a full sequence.
  arXiv  Detail & Related papers  (2025-04-23T18:38:18Z)
• Human Re-ID Meets LVLMs: What can we expect? [14.370360290704197]
  We compare the performance of the leading large vision-language models in the human re-identification task.<n>Our results confirm the strengths of LVLMs, but also their severe limitations that often lead to catastrophic answers.
  arXiv  Detail & Related papers  (2025-01-30T19:00:40Z)
• Deep Unlearn: Benchmarking Machine Unlearning for Image Classification [7.450700594277741]
  Machine unlearning (MU) aims to remove the influence of particular data points from the learnable parameters of a trained machine learning model.<n>This paper investigates 18 state-of-the-art MU methods across various benchmark datasets and models.
  arXiv  Detail & Related papers  (2024-10-02T06:41:58Z)
• Distribution-Aware Prompt Tuning for Vision-Language Models [20.02599087680773]
  A key to prompt tuning is the feature space alignment between two modalities via learnable vectors with model parameters fixed. Inspired by this observation, we proposed distribution-aware prompt tuning (DAPT) for vision-language models. Our experiments on 11 benchmark datasets demonstrate that our method significantly improves generalizability.
  arXiv  Detail & Related papers  (2023-09-06T23:49:11Z)
• Incremental Online Learning Algorithms Comparison for Gesture and Visual Smart Sensors [68.8204255655161]
  This paper compares four state-of-the-art algorithms in two real applications: gesture recognition based on accelerometer data and image classification. Our results confirm these systems' reliability and the feasibility of deploying them in tiny-memory MCUs.
  arXiv  Detail & Related papers  (2022-09-01T17:05:20Z)
• SSMTL++: Revisiting Self-Supervised Multi-Task Learning for Video Anomaly Detection [108.57862846523858]
  We revisit the self-supervised multi-task learning framework, proposing several updates to the original method. We modernize the 3D convolutional backbone by introducing multi-head self-attention modules. In our attempt to further improve the model, we study additional self-supervised learning tasks, such as predicting segmentation maps.
  arXiv  Detail & Related papers  (2022-07-16T19:25:41Z)
• Benchmarking Detection Transfer Learning with Vision Transformers [60.97703494764904]
  complexity of object detection methods can make benchmarking non-trivial when new architectures, such as Vision Transformer (ViT) models, arrive. We present training techniques that overcome these challenges, enabling the use of standard ViT models as the backbone of Mask R-CNN. Our results show that recent masking-based unsupervised learning methods may, for the first time, provide convincing transfer learning improvements on COCO.
  arXiv  Detail & Related papers  (2021-11-22T18:59:15Z)
• Efficient Training of Visual Transformers with Small-Size Datasets [64.60765211331697]
  Visual Transformers (VTs) are emerging as an architectural paradigm alternative to Convolutional networks (CNNs) We show that, despite having a comparable accuracy when trained on ImageNet, their performance on smaller datasets can be largely different. We propose a self-supervised task which can extract additional information from images with only a negligible computational overhead.
  arXiv  Detail & Related papers  (2021-06-07T16:14:06Z)

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.