paper-with-me

Papers

AttFC: Attention Fully-Connected Layer for Large-Scale Face Recognition with One GPU

2025-03-10 · Zhuowen Zheng, Yain-Whar Si, Xiaochen Yuan, Junwei Duan, Ke Wang, Xiaofan Li, Xinyuan Zhang, Xueyuan Gong

Nowadays, with the advancement of deep neural networks (DNNs) and the availability of large-scale datasets, the face recognition (FR) model has achieved exceptional performance. However, since the parameter magnitude of the fully connected (FC) layer directly depends on the number of identities in the dataset. If training the FR model on large-scale datasets, the size of the model parameter will be excessively huge, leading to substantial demand for computational resources, such as time and memory. This paper proposes the attention fully connected (AttFC) layer, which could significantly reduce computational resources. AttFC employs an attention loader to generate the generative class center (GCC), and dynamically store the class center with Dynamic Class Container (DCC). DCC only stores a small subset of all class centers in FC, thus its parameter count is substantially less than the FC layer. Also, training face recognition models on large-scale datasets with one GPU often encounter out-of-memory (OOM) issues. AttFC overcomes this and achieves comparable performance to state-of-the-art methods.

📄 PDF Abstract BibTeX arXiv:2503.06839

Code (0)

등록된 구현이 없습니다.

Tasks

Face RecognitionGPU

Methods 이 논문이 사용한 방법론

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

Similar Papers 제목 키워드 기반

DropAttention: A Regularization Method for Fully-Connected Self-Attention Networks

2019-07-25 · Lin Zehui, PengFei Liu, Luyao Huang, Junkun Chen 외

Variants dropout methods have been designed for the fully-connected layer, convolutional layer and recurrent layer in neural networks, and shown to be effective to avoid overfitting. As an appealing alternative to recurr…

Is logical analysis performed by transformers taking place in self-attention or in the fully connected part?

2025-01-20 · Evgeniy Shin, Heinrich Matzinger

Transformers architecture apply self-attention to tokens represented as vectors, before a fully connected (neuronal network) layer. These two parts can be layered many times. Traditionally, self-attention is seen as a me…

Self-Learning

Do We Need Fully Connected Output Layers in Convolutional Networks?

2020-04-28 · Zhongchao Qian, Tyler L. Hayes, Kushal Kafle, Christopher Kanan

Traditionally, deep convolutional neural networks consist of a series of convolutional and pooling layers followed by one or more fully connected (FC) layers to perform the final classification. While this design has bee…

General Classification

A Lightweight CNN-Transformer Model for Learning Traveling Salesman Problems

2023-05-03 · Minseop Jung, Jaeseung Lee, Jibum Kim

Several studies have attempted to solve traveling salesman problems (TSPs) using various deep learning techniques. Among them, Transformer-based models show state-of-the-art performance even for large-scale Traveling Sal…

GPU

Attention-based Graph Neural Network for Semi-supervised Learning

2018-03-10 · ICLR 2018 1 · Kiran K. Thekumparampil, Chong Wang, Sewoong Oh, Li-Jia Li

Recently popularized graph neural networks achieve the state-of-the-art accuracy on a number of standard benchmark datasets for graph-based semi-supervised learning, improving significantly over existing approaches. Thes…

Graph Neural NetworkGraph Regression