paper-with-me

홈 › Papers

Unveiling the Backbone-Optimizer Coupling Bias in Visual Representation Learning

2024-10-08 · Siyuan Li, Juanxi Tian, Zedong Wang, Luyuan Zhang, Zicheng Liu, Weiyang Jin, Yang Liu, Baigui Sun, Stan Z. Li

This paper delves into the interplay between vision backbones and optimizers, unvealing an inter-dependent phenomenon termed \textit{\textbf{b}ackbone-\textbf{o}ptimizer \textbf{c}oupling \textbf{b}ias} (BOCB). We observe that canonical CNNs, such as VGG and ResNet, exhibit a marked co-dependency with SGD families, while recent architectures like ViTs and ConvNeXt share a tight coupling with the adaptive learning rate ones. We further show that BOCB can be introduced by both optimizers and certain backbone designs and may significantly impact the pre-training and downstream fine-tuning of vision models. Through in-depth empirical analysis, we summarize takeaways on recommended optimizers and insights into robust vision backbone architectures. We hope this work can inspire the community to question long-held assumptions on backbones and optimizers, stimulate further explorations, and thereby contribute to more robust vision systems. The source code and models are publicly available at https://bocb-ai.github.io/.

📄 PDF Abstract BibTeX arXiv:2410.06373

Code (0)

등록된 구현이 없습니다.

Tasks

Representation Learning

Methods 이 논문이 사용한 방법론

Dropout Dropout is a regularization technique for neural networks that drops a unit (along with connections) at training time with a specified probability $p$ (a common value is…
Average Pooling 설명 없음
ConvNeXt 설명 없음
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 $x$…
Max Pooling Max Pooling is a pooling operation that calculates the maximum value for patches of a feature map, and uses it to create a downsampled (pooled) feature map. It is usually…
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…
SGD Stochastic Gradient Descent is an iterative optimization technique that uses minibatches of data to form an expectation of the gradient, rather than the full gradient using…
Kaiming Initialization 설명 없음

Similar Papers 제목 키워드 기반

Unveiling Visual Biases in Audio-Visual Localization Benchmarks

2024-08-25 · Liangyu Chen, Zihao Yue, Boshen Xu, Qin Jin

Audio-Visual Source Localization (AVSL) aims to localize the source of sound within a video. In this paper, we identify a significant issue in existing benchmarks: the sounding objects are often easily recognized based s…

audio-visual learningVisual Localization

Correcting Stochastic Update Bias in Preconditioned Language Model Optimizers

2026-05-20 · Nikhil Nayak, Julia White, Urchade Zaratiana, Kelton Zhang 외 arxiv

Preconditioned optimizers are central to language model training, but their stochastic update rules are usually treated as direct approximations to population preconditioned descent. We show that this view misses two fin…

Unveiling and Mitigating Adversarial Vulnerabilities in Iterative Optimizers

2025-04-26 · Elad Sofer, Tomer Shaked, Caroline Chaux, Nir Shlezinger

Machine learning (ML) models are often sensitive to carefully crafted yet seemingly unnoticeable perturbations. Such adversarial examples are considered to be a property of ML models, often associated with their black-bo…

Sensitivity

Two Birds, One Projection: Harmonizing Safety and Utility in LVLMs via Inference-time Feature Projection

2026-03-16 · Yewon Han, Yumin Seol, EunGyung Kong, Minsoo Jo 외 arxiv

Existing jailbreak defence frameworks for Large Vision-Language Models often suffer from a safety utility tradeoff, where strengthening safety inadvertently degrades performance on general visual-grounded reasoning tasks…

From Descriptive Richness to Bias: Unveiling the Dark Side of Generative Image Caption Enrichment

2024-06-20 · Yusuke Hirota, Ryo Hachiuma, Chao-Han Huck Yang, Yuta Nakashima

Large language models (LLMs) have enhanced the capacity of vision-language models to caption visual text. This generative approach to image caption enrichment further makes textual captions more descriptive, improving al…

DescriptiveHallucination