머신러닝 꿈나무/Paper Review

논문 링크 : https://arxiv.org/abs/2102.06171 High-Performance Large-Scale Image Recognition Without Normalization Batch normalization is a key component of most image classification models, but it has many undesirable properties stemming from its dependence on the batch size and interactions between examples. Although recent work has succeeded in training deep ResNets arxiv.org 안녕하세요? 이번에 리뷰할 논문은 NF..

논문 링크 : https://arxiv.org/abs/1801.07698 ArcFace: Additive Angular Margin Loss for Deep Face Recognition One of the main challenges in feature learning using Deep Convolutional Neural Networks (DCNNs) for large-scale face recognition is the design of appropriate loss functions that enhance discriminative power. Centre loss penalises the distance between the d arxiv.org 1500회 넘게 인용된 아주 인기있는 논문 Ar..

논문 링크 : https://arxiv.org/abs/1412.6622 Deep metric learning using Triplet network Deep learning has proven itself as a successful set of models for learning useful semantic representations of data. These, however, are mostly implicitly learned as part of a classification task. In this paper we propose the triplet network model, which ai arxiv.org 왜 이 논문을 찾아보았냐 함은... Feature Extraction을 찾다보니 Arc..

FaceNet: A unified Embedding for Face Recognition and Clustering 논문 링크 : https://arxiv.org/abs/1503.03832 FaceNet: A Unified Embedding for Face Recognition and Clustering Despite significant recent advances in the field of face recognition, implementing face verification and recognition efficiently at scale presents serious challenges to current approaches. In this paper we present a system, cal..

원문 링크 : 핵심 키워드 1. R-CNN 시리즈들의 서막 2. Regrional Proposal, Selective Search 3. Non-Maximum Suppression 4. Bounding Box Regression R-CNN 모델의 흐름 위 그림과 같이 (1)이미지를 입력하고 (2) 영역을 추출하고 (3) CNN연산을 한 후 (4)분류한다. R-CNN R-CNN이 Object Detection을 수행하는 알고리즘 입력 이미지에 selective search를 적용하여 물체가 있을만한 박스 2천개(2K)를 추출한다. 모든 박스를 227x227로 resize. 박스의 비율 등은 고려하지 않는다. imagenet으로 학습시켜놓은 CNN을 통과시켜 4096차원의 특징 벡터를 추출한다. 추출된 벡터를..