paper-with-me

Papers

Rethinking Random Masking in Self Distillation on ViT

2025-06-12 · Jihyeon Seong, Hyunkyung Han

Vision Transformers (ViTs) have demonstrated remarkable performance across a wide range of vision tasks. In particular, self-distillation frameworks such as DINO have contributed significantly to these advances. Within such frameworks, random masking is often utilized to improve training efficiency and introduce regularization. However, recent studies have raised concerns that indiscriminate random masking may inadvertently eliminate critical semantic information, motivating the development of more informed masking strategies. In this study, we explore the role of random masking in the self-distillation setting, focusing on the DINO framework. Specifically, we apply random masking exclusively to the student's global view, while preserving the student's local views and the teacher's global view in their original, unmasked forms. This design leverages DINO's multi-view augmentation scheme to retain clean supervision while inducing robustness through masked inputs. We evaluate our approach using DINO-Tiny on the mini-ImageNet dataset and show that random masking under this asymmetric setup yields more robust and fine-grained attention maps, ultimately enhancing downstream performance.

📄 PDF Abstract BibTeX arXiv:2506.10582

Code (0)

등록된 구현이 없습니다.

Methods 이 논문이 사용한 방법론

Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…
Vision Transformer The Vision Transformer, or ViT, is a model for image classification that employs a Transformer-like architecture over…
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Attention 설명 없음
DINO DINO (self-distillation with no labels) is a self-supervised learning method that directly predicts the output of a teacher network - built with a momentum encoder - using a…

Similar Papers 제목 키워드 기반

What to Hide from Your Students: Attention-Guided Masked Image Modeling

2022-03-23 · Ioannis Kakogeorgiou, Spyros Gidaris, Bill Psomas, Yannis Avrithis 외

Transformers and masked language modeling are quickly being adopted and explored in computer vision as vision transformers and masked image modeling (MIM). In this work, we argue that image token masking differs from tok…

Language ModelingLanguage ModellingMasked Language ModelingSelf-Supervised Learning

Self-supervised adversarial masking for 3D point cloud representation learning

2023-07-11 · Michał Szachniewicz, Wojciech Kozłowski, Michał Stypułkowski, Maciej Zięba

Self-supervised methods have been proven effective for learning deep representations of 3D point cloud data. Although recent methods in this domain often rely on random masking of inputs, the results of this approach can…

Representation Learning

Complementary Random Masking for RGB-Thermal Semantic Segmentation

2023-03-30 · Ukcheol Shin, Kyunghyun Lee, In So Kweon, Jean Oh

RGB-thermal semantic segmentation is one potential solution to achieve reliable semantic scene understanding in adverse weather and lighting conditions. However, the previous studies mostly focus on designing a multi-mod…

Scene UnderstandingSemantic SegmentationThermal Image Segmentation

Co-distilled attention guided masked image modeling with noisy teacher for self-supervised learning on medical images

2026-04-16 · Jue Jiang, Aneesh Rangnekar, Harini Veeraraghavan arxiv

Masked image modeling (MIM) is a highly effective self-supervised learning (SSL) approach to extract useful feature representations from unannotated data. Predominantly used random masking methods make SSL less effective…

Lung Nodule ClassificationSelf-Supervised LearningTumor Segmentation

SelFusion: Self-distillation for Diffusion Language Models

2026-08-24 · Hyeongsoo Lim, Jinyoung Kim, Eunseo Seo, Minho Jang 외 arxiv

Diffusion language models (DLMs) alleviate the inherent latency bottleneck of autoregressive (AR) large language models (LLMs), but their degraded generation quality limits practical applicability. Although knowledge dis…

Knowledge Distillation